| FFPLAY-ALL(1) | FFPLAY-ALL(1) |
ffplay - FFplay media player
ffplay [options] [input_url]
FFplay is a very simple and portable media player using the FFmpeg libraries and the SDL library. It is mostly used as a testbed for the various FFmpeg APIs.
All the numerical options, if not specified otherwise, accept a string representing a number as input, which may be followed by one of the SI unit prefixes, for example: 'K', 'M', or 'G'.
If 'i' is appended to the SI unit prefix, the complete prefix will be interpreted as a unit prefix for binary multiples, which are based on powers of 1024 instead of powers of 1000. Appending 'B' to the SI unit prefix multiplies the value by 8. This allows using, for example: 'KB', 'MiB', 'G' and 'B' as number suffixes.
Options which do not take arguments are boolean options, and set the corresponding value to true. They can be set to false by prefixing the option name with "no". For example using "-nofoo" will set the boolean option with name "foo" to false.
Some options are applied per-stream, e.g. bitrate or codec. Stream specifiers are used to precisely specify which stream(s) a given option belongs to.
A stream specifier is a string generally appended to the option name and separated from it by a colon. E.g. "-codec:a:1 ac3" contains the "a:1" stream specifier, which matches the second audio stream. Therefore, it would select the ac3 codec for the second audio stream.
A stream specifier can match several streams, so that the option is applied to all of them. E.g. the stream specifier in "-b:a 128k" matches all audio streams.
An empty stream specifier matches all streams. For example, "-codec copy" or "-codec: copy" would copy all the streams without reencoding.
Possible forms of stream specifiers are:
Note that in ffmpeg, matching by metadata will only work properly for input files.
These options are shared amongst the ff* tools.
Possible values of arg are:
Note that the term 'codec' is used throughout this documentation as a shortcut for what is more correctly called a media bitstream format.
ffmpeg -sources pulse,server=192.168.0.4
ffmpeg -sinks pulse,server=192.168.0.4
The optional flags prefix can consist of the following values:
Flags can also be used alone by adding a '+'/'-' prefix to set/reset a single flag without affecting other flags or changing loglevel. When setting both flags and loglevel, a '+' separator is expected between the last flags value and before loglevel.
loglevel is a string or a number containing one of the following values:
For example to enable repeated log output, add the "level" prefix, and set loglevel to "verbose":
ffmpeg -loglevel repeat+level+verbose -i input output
Another example that enables repeated log output without affecting current state of "level" prefix flag or loglevel:
ffmpeg [...] -loglevel +repeat
By default the program logs to stderr. If coloring is supported by the terminal, colors are used to mark errors and warnings. Log coloring can be disabled setting the environment variable AV_LOG_FORCE_NOCOLOR, or can be forced setting the environment variable AV_LOG_FORCE_COLOR.
Setting the environment variable FFREPORT to any value has the same effect. If the value is a ':'-separated key=value sequence, these options will affect the report; option values must be escaped if they contain special characters or the options delimiter ':' (see the ``Quoting and escaping'' section in the ffmpeg-utils manual).
The following options are recognized:
For example, to output a report to a file named ffreport.log using a log level of 32 (alias for log level "info"):
FFREPORT=file=ffreport.log:level=32 ffmpeg -i input output
Errors in parsing the environment variable are not fatal, and will not appear in the report.
All FFmpeg tools will normally show a copyright notice, build options and library versions. This option can be used to suppress printing this information.
ffmpeg -cpuflags -sse+mmx ...
ffmpeg -cpuflags mmx ...
ffmpeg -cpuflags 0 ...
Possible flags for this option are:
ffmpeg -cpucount 2
These options are provided directly by the libavformat, libavdevice and libavcodec libraries. To see the list of available AVOptions, use the -help option. They are separated into two categories:
For example to write an ID3v2.3 header instead of a default ID3v2.4 to an MP3 file, use the id3v2_version private option of the MP3 muxer:
ffmpeg -i input.flac -id3v2_version 3 out.mp3
All codec AVOptions are per-stream, and thus a stream specifier should be attached to them:
ffmpeg -i multichannel.mxf -map 0:v:0 -map 0:a:0 -map 0:a:0 -c:a:0 ac3 -b:a:0 640k -ac:a:1 2 -c:a:1 aac -b:2 128k out.mp4
In the above example, a multichannel audio stream is mapped twice for output. The first instance is encoded with codec ac3 and bitrate 640k. The second instance is downmixed to 2 channels and encoded with codec aac. A bitrate of 128k is specified for it using absolute index of the output stream.
Note: the -nooption syntax cannot be used for boolean AVOptions, use -option 0/-option 1.
Note: the old undocumented way of specifying per-stream AVOptions by prepending v/a/s to the options name is now obsolete and will be removed soon.
pos must be a time duration specification, see the Time duration section in the ffmpeg-utils(1) manual.
duration must be a time duration specification, see the Time duration section in the ffmpeg-utils(1) manual.
Default value is "video", if video is not present or cannot be played "rdft" is automatically selected.
You can interactively cycle through the available show modes by pressing the key w.
filtergraph is a description of the filtergraph to apply to the stream, and must have a single video input and a single video output. In the filtergraph, the input is associated to the label "in", and the output to the label "out". See the ffmpeg-filters manual for more information about the filtergraph syntax.
You can specify this parameter multiple times and cycle through the specified filtergraphs along with the show modes by pressing the key w.
Pause if the stream is not already paused, step to the next video frame, and pause.
This section documents the syntax and formats employed by the FFmpeg libraries and tools.
FFmpeg adopts the following quoting and escaping mechanism, unless explicitly specified. The following rules are applied:
Note that you may need to add a second level of escaping when using the command line or a script, which depends on the syntax of the adopted shell language.
The function "av_get_token" defined in libavutil/avstring.h can be used to parse a token quoted or escaped according to the rules defined above.
The tool tools/ffescape in the FFmpeg source tree can be used to automatically quote or escape a string in a script.
Examples
Crime d\'Amour
'Crime d'\''Amour'
' this string starts and ends with whitespaces '
' The string '\'string\'' is a string '
'c:\foo' can be written as c:\\foo
The accepted syntax is:
[(YYYY-MM-DD|YYYYMMDD)[T|t| ]]((HH:MM:SS[.m...]]])|(HHMMSS[.m...]]]))[Z]
now
If the value is "now" it takes the current time.
Time is local time unless Z is appended, in which case it is interpreted as UTC. If the year-month-day part is not specified it takes the current year-month-day.
There are two accepted syntaxes for expressing time duration.
[-][<HH>:]<MM>:<SS>[.<m>...]
HH expresses the number of hours, MM the number of minutes for a maximum of 2 digits, and SS the number of seconds for a maximum of 2 digits. The m at the end expresses decimal value for SS.
or
[-]<S>+[.<m>...][s|ms|us]
S expresses the number of seconds, with the optional decimal part m. The optional literal suffixes s, ms or us indicate to interpret the value as seconds, milliseconds or microseconds, respectively.
In both expressions, the optional - indicates negative duration.
Examples
The following examples are all valid time duration:
Specify the size of the sourced video, it may be a string of the form widthxheight, or the name of a size abbreviation.
The following abbreviations are recognized:
Specify the frame rate of a video, expressed as the number of frames generated per second. It has to be a string in the format frame_rate_num/frame_rate_den, an integer number, a float number or a valid video frame rate abbreviation.
The following abbreviations are recognized:
A ratio can be expressed as an expression, or in the form numerator:denominator.
Note that a ratio with infinite (1/0) or negative value is considered valid, so you should check on the returned value if you want to exclude those values.
The undefined value can be expressed using the "0:0" string.
It can be the name of a color as defined below (case insensitive match) or a "[0x|#]RRGGBB[AA]" sequence, possibly followed by @ and a string representing the alpha component.
The alpha component may be a string composed by "0x" followed by an hexadecimal number or a decimal number between 0.0 and 1.0, which represents the opacity value (0x00 or 0.0 means completely transparent, 0xff or 1.0 completely opaque). If the alpha component is not specified then 0xff is assumed.
The string random will result in a random color.
The following names of colors are recognized:
A channel layout specifies the spatial disposition of the channels in a multi-channel audio stream. To specify a channel layout, FFmpeg makes use of a special syntax.
Individual channels are identified by an id, as given by the table below:
Standard channel layout compositions can be specified by using the following identifiers:
A custom channel layout can be specified as a sequence of terms, separated by '+'. Each term can be:
A standard channel layout can be specified by the following:
Before libavutil version 53 the trailing character "c" to specify a number of channels was optional, but now it is required, while a channel layout mask can also be specified as a decimal number (if and only if not followed by "c" or "C").
See also the function "av_channel_layout_from_string" defined in libavutil/channel_layout.h.
When evaluating an arithmetic expression, FFmpeg uses an internal formula evaluator, implemented through the libavutil/eval.h interface.
An expression may contain unary, binary operators, constants, and functions.
Two expressions expr1 and expr2 can be combined to form another expression "expr1;expr2". expr1 and expr2 are evaluated in turn, and the new expression evaluates to the value of expr2.
The following binary operators are available: "+", "-", "*", "/", "^".
The following unary operators are available: "+", "-".
The following functions are available:
The results of the evaluation of x and y are converted to integers before executing the bitwise operation.
Note that both the conversion to integer and the conversion back to floating point can lose precision. Beware of unexpected results for large numbers (usually 2^53 and larger).
Prints t with loglevel l
The expression in expr must denote a continuous function or the result is undefined.
ld(0) is used to represent the function input value, which means that the given expression will be evaluated multiple times with various input values that the expression can access through ld(0). When the expression evaluates to 0 then the corresponding input value will be returned.
When the series does not converge the result is undefined.
ld(id) is used to represent the derivative order in expr, which means that the given expression will be evaluated multiple times with various input values that the expression can access through ld(id). If id is not specified then 0 is assumed.
Note, when you have the derivatives at y instead of 0, "taylor(expr, x-y)" can be used.
The following constants are available:
Assuming that an expression is considered "true" if it has a non-zero value, note that:
"*" works like AND
"+" works like OR
For example the construct:
if (A AND B) then C
is equivalent to:
if(A*B, C)
In your C code, you can extend the list of unary and binary functions, and define recognized constants, so that they are available for your expressions.
The evaluator also recognizes the International System unit prefixes. If 'i' is appended after the prefix, binary prefixes are used, which are based on powers of 1024 instead of powers of 1000. The 'B' postfix multiplies the value by 8, and can be appended after a unit prefix or used alone. This allows using for example 'KB', 'MiB', 'G' and 'B' as number postfix.
The list of available International System prefixes follows, with indication of the corresponding powers of 10 and of 2.
libavcodec provides some generic global options, which can be set on all the encoders and decoders. In addition each codec may support so-called private options, which are specific for a given codec.
Sometimes, a global option may only affect a specific kind of codec, and may be nonsensical or ignored by another, so you need to be aware of the meaning of the specified options. Also some options are meant only for decoding or encoding.
Options may be set by specifying -option value in the FFmpeg tools, or by setting the value explicitly in the "AVCodecContext" options or using the libavutil/opt.h API for programmatic use.
The list of supported options follow:
Possible values:
It is the fundamental unit of time (in seconds) in terms of which frame timestamps are represented. For fixed-fps content, timebase should be "1 / frame_rate" and timestamp increments should be identically 1.
Each submitted frame except the last must contain exactly frame_size samples per channel. May be 0 when the codec has CODEC_CAP_VARIABLE_FRAME_SIZE set, in that case the frame size is not restricted. It is set by some decoders to indicate constant frame size.
Must be an integer between -1 and 16. 0 means that B-frames are disabled. If a value of -1 is used, it will choose an automatic value depending on the encoder.
Default value is 0.
Possible values:
Possible values:
Possible values:
Possible values:
Possible values:
Possible values:
Possible values:
Possible values:
Possible values:
Possible values:
Possible values:
Possible values:
Possible values:
Possible values:
Default value is auto.
Possible values:
skip_loop_filter skips frame loop filtering, skip_idct skips frame IDCT/dequantization, skip_frame skips decoding.
Possible values:
Default value is default.
Use of frame will increase decoding delay by one frame per thread, so clients which cannot provide future frames should not use it.
Possible values:
Default value is slice+frame.
Possible values:
ffprobe -dump_separator "
" -i ~/videos/matrixbench_mpeg2.mpg
Decoders are configured elements in FFmpeg which allow the decoding of multimedia streams.
When you configure your FFmpeg build, all the supported native decoders are enabled by default. Decoders requiring an external library must be enabled manually via the corresponding "--enable-lib" option. You can list all available decoders using the configure option "--list-decoders".
You can disable all the decoders with the configure option "--disable-decoders" and selectively enable / disable single decoders with the options "--enable-decoder=DECODER" / "--disable-decoder=DECODER".
The option "-decoders" of the ff* tools will display the list of enabled decoders.
A description of some of the currently available video decoders follows.
AOMedia Video 1 (AV1) decoder.
Options
Raw video decoder.
This decoder decodes rawvideo streams.
Options
dav1d AV1 decoder.
libdav1d allows libavcodec to decode the AOMedia Video 1 (AV1) codec. Requires the presence of the libdav1d headers and library during configuration. You need to explicitly configure the build with "--enable-libdav1d".
Options
The following options are supported by the libdav1d wrapper.
AVS2-P2/IEEE1857.4 video decoder wrapper.
This decoder allows libavcodec to decode AVS2 streams with davs2 library.
AVS3-P2/IEEE1857.10 video decoder.
libuavs3d allows libavcodec to decode AVS3 streams. Requires the presence of the libuavs3d headers and library during configuration. You need to explicitly configure the build with "--enable-libuavs3d".
Options
The following option is supported by the libuavs3d wrapper.
The family of Intel QuickSync Video decoders (VC1, MPEG-2, H.264, HEVC, JPEG/MJPEG, VP8, VP9, AV1).
Common Options
The following options are supported by all qsv decoders.
HEVC Options
Extra options for hevc_qsv.
Uncompressed 4:2:2 10-bit decoder.
Options
A description of some of the currently available audio decoders follows.
AC-3 audio decoder.
This decoder implements part of ATSC A/52:2010 and ETSI TS 102 366, as well as the undocumented RealAudio 3 (a.k.a. dnet).
AC-3 Decoder Options
FLAC audio decoder.
This decoder aims to implement the complete FLAC specification from Xiph.
FLAC Decoder options
Internal wave synthesizer.
This decoder generates wave patterns according to predefined sequences. Its use is purely internal and the format of the data it accepts is not publicly documented.
libcelt decoder wrapper.
libcelt allows libavcodec to decode the Xiph CELT ultra-low delay audio codec. Requires the presence of the libcelt headers and library during configuration. You need to explicitly configure the build with "--enable-libcelt".
libgsm decoder wrapper.
libgsm allows libavcodec to decode the GSM full rate audio codec. Requires the presence of the libgsm headers and library during configuration. You need to explicitly configure the build with "--enable-libgsm".
This decoder supports both the ordinary GSM and the Microsoft variant.
libilbc decoder wrapper.
libilbc allows libavcodec to decode the Internet Low Bitrate Codec (iLBC) audio codec. Requires the presence of the libilbc headers and library during configuration. You need to explicitly configure the build with "--enable-libilbc".
Options
The following option is supported by the libilbc wrapper.
libopencore-amrnb decoder wrapper.
libopencore-amrnb allows libavcodec to decode the Adaptive Multi-Rate Narrowband audio codec. Using it requires the presence of the libopencore-amrnb headers and library during configuration. You need to explicitly configure the build with "--enable-libopencore-amrnb".
An FFmpeg native decoder for AMR-NB exists, so users can decode AMR-NB without this library.
libopencore-amrwb decoder wrapper.
libopencore-amrwb allows libavcodec to decode the Adaptive Multi-Rate Wideband audio codec. Using it requires the presence of the libopencore-amrwb headers and library during configuration. You need to explicitly configure the build with "--enable-libopencore-amrwb".
An FFmpeg native decoder for AMR-WB exists, so users can decode AMR-WB without this library.
libopus decoder wrapper.
libopus allows libavcodec to decode the Opus Interactive Audio Codec. Requires the presence of the libopus headers and library during configuration. You need to explicitly configure the build with "--enable-libopus".
An FFmpeg native decoder for Opus exists, so users can decode Opus without this library.
ARIB STD-B24 caption decoder.
Implements profiles A and C of the ARIB STD-B24 standard.
libaribb24 Decoder Options
Unset by default.
Enabled by default.
Yet another ARIB STD-B24 caption decoder using external libaribcaption library.
Implements profiles A and C of the Japanse ARIB STD-B24 standard, Brazilian ABNT NBR 15606-1, and Philippines version of ISDB-T.
Requires the presence of the libaribcaption headers and library (<https://github.com/xqq/libaribcaption>) during configuration. You need to explicitly configure the build with "--enable-libaribcaption". If both libaribb24 and libaribcaption are enabled, libaribcaption decoder precedes.
libaribcaption Decoder Options
The default is ass as same as libaribb24 decoder. Some present players (e.g., mpv) expect ASS format for ARIB caption.
If not specified, use internaly defined default font family.
The default is false.
If your player cannot handle AVSubtitles with multiple ASS rectangles properly, set this option to true or define ASS_SINGLE_RECT=1 to change default behavior at compilation.
The default is false.
The default is 1.5.
The default is false.
The default is false.
The default is true.
The default is true.
The default is true.
The default is true.
The libaribcaption decoder assumes input frame size for bitmap rendering as below:
If actual frame size of input video does not match above assumption, the rendered captions may be distorted. To make the captions undistorted, add "-canvas_size" option to specify actual input video size.
Note that the "-canvas_size" option is not required for video with different size but same aspect ratio. In such cases, the caption will be stretched or shrunk to actual video size if "-canvas_size" option is not specified. If "-canvas_size" option is specified with different size, the caption will be stretched or shrunk as specified size with calculated SAR.
libaribcaption decoder usage examples
Display MPEG-TS file with ARIB subtitle by "ffplay" tool:
ffplay -sub_type bitmap MPEG.TS
Display MPEG-TS file with input frame size 1920x1080 by "ffplay" tool:
ffplay -sub_type bitmap -canvas_size 1920x1080 MPEG.TS
Embed ARIB subtitle in transcoded video:
ffmpeg -sub_type bitmap -i src.m2t -filter_complex "[0:v][0:s]overlay" -vcodec h264 dest.mp4
Options
This codec decodes the bitmap subtitles used in DVDs; the same subtitles can also be found in VobSub file pairs and in some Matroska files.
Options
The format for this option is a string containing 16 24-bits hexadecimal numbers (without 0x prefix) separated by commas, for example "0d00ee, ee450d, 101010, eaeaea, 0ce60b, ec14ed, ebff0b, 0d617a, 7b7b7b, d1d1d1, 7b2a0e, 0d950c, 0f007b, cf0dec, cfa80c, 7c127b".
Libzvbi allows libavcodec to decode DVB teletext pages and DVB teletext subtitles. Requires the presence of the libzvbi headers and library during configuration. You need to explicitly configure the build with "--enable-libzvbi".
Options
When you configure your FFmpeg build, all the supported bitstream filters are enabled by default. You can list all available ones using the configure option "--list-bsfs".
You can disable all the bitstream filters using the configure option "--disable-bsfs", and selectively enable any bitstream filter using the option "--enable-bsf=BSF", or you can disable a particular bitstream filter using the option "--disable-bsf=BSF".
The option "-bsfs" of the ff* tools will display the list of all the supported bitstream filters included in your build.
The ff* tools have a -bsf option applied per stream, taking a comma-separated list of filters, whose parameters follow the filter name after a '='.
ffmpeg -i INPUT -c:v copy -bsf:v filter1[=opt1=str1:opt2=str2][,filter2] OUTPUT
Below is a description of the currently available bitstream filters, with their parameters, if any.
Convert MPEG-2/4 AAC ADTS to an MPEG-4 Audio Specific Configuration bitstream.
This filter creates an MPEG-4 AudioSpecificConfig from an MPEG-2/4 ADTS header and removes the ADTS header.
This filter is required for example when copying an AAC stream from a raw ADTS AAC or an MPEG-TS container to MP4A-LATM, to an FLV file, or to MOV/MP4 files and related formats such as 3GP or M4A. Please note that it is auto-inserted for MP4A-LATM and MOV/MP4 and related formats.
Modify metadata embedded in an AV1 stream.
Remove zero padding at the end of a packet.
Extract the core from a DCA/DTS stream, dropping extensions such as DTS-HD.
Add extradata to the beginning of the filtered packets except when said packets already exactly begin with the extradata that is intended to be added.
If not specified it is assumed k.
For example the following ffmpeg command forces a global header (thus disabling individual packet headers) in the H.264 packets generated by the "libx264" encoder, but corrects them by adding the header stored in extradata to the key packets:
ffmpeg -i INPUT -map 0 -flags:v +global_header -c:v libx264 -bsf:v dump_extra out.ts
Blocks in DV which are marked as damaged are replaced by blocks of the specified color.
see page 44-46 or section 5.5 of <http://web.archive.org/web/20060927044735/http://www.smpte.org/smpte_store/standards/pdf/s314m.pdf>
Extract the core from a E-AC-3 stream, dropping extra channels.
Extract the in-band extradata.
Certain codecs allow the long-term headers (e.g. MPEG-2 sequence headers, or H.264/HEVC (VPS/)SPS/PPS) to be transmitted either "in-band" (i.e. as a part of the bitstream containing the coded frames) or "out of band" (e.g. on the container level). This latter form is called "extradata" in FFmpeg terminology.
This bitstream filter detects the in-band headers and makes them available as extradata.
Remove units with types in or not in a given set from the stream.
Extradata is unchanged by this transformation, but note that if the stream contains inline parameter sets then the output may be unusable if they are removed.
For example, to remove all non-VCL NAL units from an H.264 stream:
ffmpeg -i INPUT -c:v copy -bsf:v 'filter_units=pass_types=1-5' OUTPUT
To remove all AUDs, SEI and filler from an H.265 stream:
ffmpeg -i INPUT -c:v copy -bsf:v 'filter_units=remove_types=35|38-40' OUTPUT
Extract Rgb or Alpha part of an HAPQA file, without recompression, in order to create an HAPQ or an HAPAlphaOnly file.
Convert HAPQA to HAPQ
ffmpeg -i hapqa_inputfile.mov -c copy -bsf:v hapqa_extract=texture=color -tag:v HapY -metadata:s:v:0 encoder="HAPQ" hapq_file.mov
Convert HAPQA to HAPAlphaOnly
ffmpeg -i hapqa_inputfile.mov -c copy -bsf:v hapqa_extract=texture=alpha -tag:v HapA -metadata:s:v:0 encoder="HAPAlpha Only" hapalphaonly_file.mov
Modify metadata embedded in an H.264 stream.
Default is pass.
These fields are set in pixels. Note that some sizes may not be representable if the chroma is subsampled or the stream is interlaced (see H.264 section 7.4.2.1.1).
For example, 086f3693-b7b3-4f2c-9653-21492feee5b8+hello will insert the string ``hello'' associated with the given UUID.
Default is pass.
Insert mode works in conjunction with "rotate" and "flip" options. Any pre-existing Display orientation messages will be removed in insert or remove mode. Extract mode attaches the display matrix to the packet as side data.
Default is unset.
The argument must be the name of a level (for example, 4.2), a level_idc value (for example, 42), or the special name auto indicating that the filter should attempt to guess the level from the input stream properties.
Convert an H.264 bitstream from length prefixed mode to start code prefixed mode (as defined in the Annex B of the ITU-T H.264 specification).
This is required by some streaming formats, typically the MPEG-2 transport stream format (muxer "mpegts").
For example to remux an MP4 file containing an H.264 stream to mpegts format with ffmpeg, you can use the command:
ffmpeg -i INPUT.mp4 -codec copy -bsf:v h264_mp4toannexb OUTPUT.ts
Please note that this filter is auto-inserted for MPEG-TS (muxer "mpegts") and raw H.264 (muxer "h264") output formats.
This applies a specific fixup to some Blu-ray streams which contain redundant PPSs modifying irrelevant parameters of the stream which confuse other transformations which require correct extradata.
Modify metadata embedded in an HEVC stream.
These fields are set in pixels. Note that some sizes may not be representable if the chroma is subsampled (H.265 section 7.4.3.2.1).
The argument must be the name of a level (for example, 5.1), a general_level_idc value (for example, 153 for level 5.1), or the special name auto indicating that the filter should attempt to guess the level from the input stream properties.
Convert an HEVC/H.265 bitstream from length prefixed mode to start code prefixed mode (as defined in the Annex B of the ITU-T H.265 specification).
This is required by some streaming formats, typically the MPEG-2 transport stream format (muxer "mpegts").
For example to remux an MP4 file containing an HEVC stream to mpegts format with ffmpeg, you can use the command:
ffmpeg -i INPUT.mp4 -codec copy -bsf:v hevc_mp4toannexb OUTPUT.ts
Please note that this filter is auto-inserted for MPEG-TS (muxer "mpegts") and raw HEVC/H.265 (muxer "h265" or "hevc") output formats.
Modifies the bitstream to fit in MOV and to be usable by the Final Cut Pro decoder. This filter only applies to the mpeg2video codec, and is likely not needed for Final Cut Pro 7 and newer with the appropriate -tag:v.
For example, to remux 30 MB/sec NTSC IMX to MOV:
ffmpeg -i input.mxf -c copy -bsf:v imxdump -tag:v mx3n output.mov
Convert MJPEG/AVI1 packets to full JPEG/JFIF packets.
MJPEG is a video codec wherein each video frame is essentially a JPEG image. The individual frames can be extracted without loss, e.g. by
ffmpeg -i ../some_mjpeg.avi -c:v copy frames_%d.jpg
Unfortunately, these chunks are incomplete JPEG images, because they lack the DHT segment required for decoding. Quoting from <http://www.digitalpreservation.gov/formats/fdd/fdd000063.shtml>:
Avery Lee, writing in the rec.video.desktop newsgroup in 2001, commented that "MJPEG, or at least the MJPEG in AVIs having the MJPG fourcc, is restricted JPEG with a fixed -- and *omitted* -- Huffman table. The JPEG must be YCbCr colorspace, it must be 4:2:2, and it must use basic Huffman encoding, not arithmetic or progressive. . . . You can indeed extract the MJPEG frames and decode them with a regular JPEG decoder, but you have to prepend the DHT segment to them, or else the decoder won't have any idea how to decompress the data. The exact table necessary is given in the OpenDML spec."
This bitstream filter patches the header of frames extracted from an MJPEG stream (carrying the AVI1 header ID and lacking a DHT segment) to produce fully qualified JPEG images.
ffmpeg -i mjpeg-movie.avi -c:v copy -bsf:v mjpeg2jpeg frame_%d.jpg
exiftran -i -9 frame*.jpg
ffmpeg -i frame_%d.jpg -c:v copy rotated.avi
Add an MJPEG A header to the bitstream, to enable decoding by Quicktime.
Extract a representable text file from MOV subtitles, stripping the metadata header from each subtitle packet.
See also the text2movsub filter.
Decompress non-standard compressed MP3 audio headers.
Modify metadata embedded in an MPEG-2 stream.
The following fixed values are supported:
Any other value will result in square pixels being signalled instead (see H.262 section 6.3.3 and table 6-3).
Unpack DivX-style packed B-frames.
DivX-style packed B-frames are not valid MPEG-4 and were only a workaround for the broken Video for Windows subsystem. They use more space, can cause minor AV sync issues, require more CPU power to decode (unless the player has some decoded picture queue to compensate the 2,0,2,0 frame per packet style) and cause trouble if copied into a standard container like mp4 or mpeg-ps/ts, because MPEG-4 decoders may not be able to decode them, since they are not valid MPEG-4.
For example to fix an AVI file containing an MPEG-4 stream with DivX-style packed B-frames using ffmpeg, you can use the command:
ffmpeg -i INPUT.avi -codec copy -bsf:v mpeg4_unpack_bframes OUTPUT.avi
Damages the contents of packets or simply drops them without damaging the container. Can be used for fuzzing or testing error resilience/concealment.
Parameters:
Both "amount" and "drop" accept expressions containing the following variables:
Examples
Apply modification to every byte but don't drop any packets.
ffmpeg -i INPUT -c copy -bsf noise=1 output.mkv
Drop every video packet not marked as a keyframe after timestamp 30s but do not modify any of the remaining packets.
ffmpeg -i INPUT -c copy -bsf:v noise=drop='gt(t\,30)*not(key)' output.mkv
Drop one second of audio every 10 seconds and add some random noise to the rest.
ffmpeg -i INPUT -c copy -bsf:a noise=amount=-1:drop='between(mod(t\,10)\,9\,10)' output.mkv
This bitstream filter passes the packets through unchanged.
Repacketize PCM audio to a fixed number of samples per packet or a fixed packet rate per second. This is similar to the asetnsamples audio filter but works on audio packets instead of audio frames.
You can generate the well known 1602-1601-1602-1601-1602 pattern of 48kHz audio for NTSC frame rate using the frame_rate option.
ffmpeg -f lavfi -i sine=r=48000:d=1 -c pcm_s16le -bsf pcm_rechunk=r=30000/1001 -f framecrc -
Merge a sequence of PGS Subtitle segments ending with an "end of display set" segment into a single packet.
This is required by some containers that support PGS subtitles (muxer "matroska").
Modify color property metadata embedded in prores stream.
Set Rec709 colorspace for each frame of the file
ffmpeg -i INPUT -c copy -bsf:v prores_metadata=color_primaries=bt709:color_trc=bt709:colorspace=bt709 output.mov
Set Hybrid Log-Gamma parameters for each frame of the file
ffmpeg -i INPUT -c copy -bsf:v prores_metadata=color_primaries=bt2020:color_trc=arib-std-b67:colorspace=bt2020nc output.mov
Remove extradata from packets.
It accepts the following parameter:
Set PTS and DTS in packets.
It accepts the following parameters:
The expressions are evaluated through the eval API and can contain the following constants:
Convert text subtitles to MOV subtitles (as used by the "mov_text" codec) with metadata headers.
See also the mov2textsub filter.
Log trace output containing all syntax elements in the coded stream headers (everything above the level of individual coded blocks). This can be useful for debugging low-level stream issues.
Supports AV1, H.264, H.265, (M)JPEG, MPEG-2 and VP9, but depending on the build only a subset of these may be available.
Extract the core from a TrueHD stream, dropping ATMOS data.
Modify metadata embedded in a VP9 stream.
Merge VP9 invisible (alt-ref) frames back into VP9 superframes. This fixes merging of split/segmented VP9 streams where the alt-ref frame was split from its visible counterpart.
Split VP9 superframes into single frames.
Given a VP9 stream with correct timestamps but possibly out of order, insert additional show-existing-frame packets to correct the ordering.
The libavformat library provides some generic global options, which can be set on all the muxers and demuxers. In addition each muxer or demuxer may support so-called private options, which are specific for that component.
Options may be set by specifying -option value in the FFmpeg tools, or by setting the value explicitly in the "AVFormatContext" options or using the libavutil/opt.h API for programmatic use.
The list of supported options follows:
Possible values for input files:
Possible values for output files:
Possible values:
Possible values:
To ensure all the streams are interleaved correctly, libavformat will wait until it has at least one packet for each stream before actually writing any packets to the output file. When some streams are "sparse" (i.e. there are large gaps between successive packets), this can result in excessive buffering.
This field specifies the maximum difference between the timestamps of the first and the last packet in the muxing queue, above which libavformat will output a packet regardless of whether it has queued a packet for all the streams.
If set to 0, libavformat will continue buffering packets until it has a packet for each stream, regardless of the maximum timestamp difference between the buffered packets.
When shifting is enabled, all output timestamps are shifted by the same amount. Audio, video, and subtitles desynching and relative timestamp differences are preserved compared to how they would have been without shifting.
offset must be a time duration specification, see the Time duration section in the ffmpeg-utils(1) manual.
The offset is added by the muxer to the output timestamps.
Specifying a positive offset means that the corresponding streams are delayed bt the time duration specified in offset. Default value is 0 (meaning that no offset is applied).
ffprobe -dump_separator "
" -i ~/videos/matrixbench_mpeg2.mpg
Possible values:
Format stream specifiers allow selection of one or more streams that match specific properties.
The exact semantics of stream specifiers is defined by the avformat_match_stream_specifier() function declared in the libavformat/avformat.h header and documented in the Stream specifiers section in the ffmpeg(1) manual.
Demuxers are configured elements in FFmpeg that can read the multimedia streams from a particular type of file.
When you configure your FFmpeg build, all the supported demuxers are enabled by default. You can list all available ones using the configure option "--list-demuxers".
You can disable all the demuxers using the configure option "--disable-demuxers", and selectively enable a single demuxer with the option "--enable-demuxer=DEMUXER", or disable it with the option "--disable-demuxer=DEMUXER".
The option "-demuxers" of the ff* tools will display the list of enabled demuxers. Use "-formats" to view a combined list of enabled demuxers and muxers.
The description of some of the currently available demuxers follows.
Audible Format 2, 3, and 4 demuxer.
This demuxer is used to demux Audible Format 2, 3, and 4 (.aa) files.
Raw Audio Data Transport Stream AAC demuxer.
This demuxer is used to demux an ADTS input containing a single AAC stream alongwith any ID3v1/2 or APE tags in it.
Animated Portable Network Graphics demuxer.
This demuxer is used to demux APNG files. All headers, but the PNG signature, up to (but not including) the first fcTL chunk are transmitted as extradata. Frames are then split as being all the chunks between two fcTL ones, or between the last fcTL and IEND chunks.
Advanced Systems Format demuxer.
This demuxer is used to demux ASF files and MMS network streams.
Virtual concatenation script demuxer.
This demuxer reads a list of files and other directives from a text file and demuxes them one after the other, as if all their packets had been muxed together.
The timestamps in the files are adjusted so that the first file starts at 0 and each next file starts where the previous one finishes. Note that it is done globally and may cause gaps if all streams do not have exactly the same length.
All files must have the same streams (same codecs, same time base, etc.).
The duration of each file is used to adjust the timestamps of the next file: if the duration is incorrect (because it was computed using the bit-rate or because the file is truncated, for example), it can cause artifacts. The "duration" directive can be used to override the duration stored in each file.
Syntax
The script is a text file in extended-ASCII, with one directive per line. Empty lines, leading spaces and lines starting with '#' are ignored. The following directive is recognized:
All subsequent file-related directives apply to that file.
To make FFmpeg recognize the format automatically, this directive must appear exactly as is (no extra space or byte-order-mark) on the very first line of the script.
If the duration is set for all files, then it is possible to seek in the whole concatenated video.
This directive works best with intra frame codecs, because for non-intra frame ones you will usually get extra packets before the actual In point and the decoded content will most likely contain frames before In point too.
For each file, packets before the file In point will have timestamps less than the calculated start timestamp of the file (negative in case of the first file), and the duration of the files (if not specified by the "duration" directive) will be reduced based on their specified In point.
Because of potential packets before the specified In point, packet timestamps may overlap between two concatenated files.
Out point is exclusive, which means that the demuxer will not output packets with a decoding timestamp greater or equal to Out point.
This directive works best with intra frame codecs and formats where all streams are tightly interleaved. For non-intra frame codecs you will usually get additional packets with presentation timestamp after Out point therefore the decoded content will most likely contain frames after Out point too. If your streams are not tightly interleaved you may not get all the packets from all streams before Out point and you may only will be able to decode the earliest stream until Out point.
The duration of the files (if not specified by the "duration" directive) will be reduced based on their specified Out point.
Options
This demuxer accepts the following option:
If set to 0, any file name is accepted.
The default is 1.
Currently, the only conversion is adding the h264_mp4toannexb bitstream filter to H.264 streams in MP4 format. This is necessary in particular if there are resolution changes.
Examples
# my first filename
file /mnt/share/file-1.wav
# my second filename including whitespace
file '/mnt/share/file 2.wav'
# my third filename including whitespace plus single quote
file '/mnt/share/file 3'\''.wav'
ffconcat version 1.0
file file-1.wav
duration 20.0
file subdir/file-2.wav
Dynamic Adaptive Streaming over HTTP demuxer.
This demuxer presents all AVStreams found in the manifest. By setting the discard flags on AVStreams the caller can decide which streams to actually receive. Each stream mirrors the "id" and "bandwidth" properties from the "<Representation>" as metadata keys named "id" and "variant_bitrate" respectively.
Options
This demuxer accepts the following option:
Electronic Arts Multimedia format demuxer.
This format is used by various Electronic Arts games.
Options
Interoperable Master Format demuxer.
This demuxer presents audio and video streams found in an IMF Composition, as specified in <https://doi.org/10.5594/SMPTE.ST2067-2.2020>.
ffmpeg [-assetmaps <path of ASSETMAP1>,<path of ASSETMAP2>,...] -i <path of CPL> ...
If "-assetmaps" is not specified, the demuxer looks for a file called ASSETMAP.xml in the same directory as the CPL.
Adobe Flash Video Format demuxer.
This demuxer is used to demux FLV files and RTMP network streams. In case of live network streams, if you force format, you may use live_flv option instead of flv to survive timestamp discontinuities. KUX is a flv variant used on the Youku platform.
ffmpeg -f flv -i myfile.flv ...
ffmpeg -f live_flv -i rtmp://<any.server>/anything/key ....
Animated GIF demuxer.
It accepts the following options:
For example, with the overlay filter, place an infinitely looping GIF over another video:
ffmpeg -i input.mp4 -ignore_loop 0 -i input.gif -filter_complex overlay=shortest=1 out.mkv
Note that in the above example the shortest option for overlay filter is used to end the output video at the length of the shortest input file, which in this case is input.mp4 as the GIF in this example loops infinitely.
HLS demuxer
Apple HTTP Live Streaming demuxer.
This demuxer presents all AVStreams from all variant streams. The id field is set to the bitrate variant index number. By setting the discard flags on AVStreams (by pressing 'a' or 'v' in ffplay), the caller can decide which variant streams to actually receive. The total bitrate of the variant that the stream belongs to is available in a metadata key named "variant_bitrate".
It accepts the following options:
Image file demuxer.
This demuxer reads from a list of image files specified by a pattern. The syntax and meaning of the pattern is specified by the option pattern_type.
The pattern may contain a suffix which is used to automatically determine the format of the images contained in the files.
The size, the pixel format, and the format of each image must be the same for all the files in the sequence.
This demuxer accepts the following options:
pattern_type accepts one of the following values.
A sequence pattern may contain the string "%d" or "%0Nd", which specifies the position of the characters representing a sequential number in each filename matched by the pattern. If the form "%d0Nd" is used, the string representing the number in each filename is 0-padded and N is the total number of 0-padded digits representing the number. The literal character '%' can be specified in the pattern with the string "%%".
If the sequence pattern contains "%d" or "%0Nd", the first filename of the file list specified by the pattern must contain a number inclusively contained between start_number and start_number+start_number_range-1, and all the following numbers must be sequential.
For example the pattern "img-%03d.bmp" will match a sequence of filenames of the form img-001.bmp, img-002.bmp, ..., img-010.bmp, etc.; the pattern "i%%m%%g-%d.jpg" will match a sequence of filenames of the form i%m%g-1.jpg, i%m%g-2.jpg, ..., i%m%g-10.jpg, etc.
Note that the pattern must not necessarily contain "%d" or "%0Nd", for example to convert a single image file img.jpeg you can employ the command:
ffmpeg -i img.jpeg img.png
The pattern is interpreted like a glob() pattern. This is only selectable if libavformat was compiled with globbing support.
If your version of libavformat was compiled with globbing support, and the provided pattern contains at least one glob meta character among "%*?[]{}" that is preceded by an unescaped "%", the pattern is interpreted like a glob() pattern, otherwise it is interpreted like a sequence pattern.
All glob special characters "%*?[]{}" must be prefixed with "%". To escape a literal "%" you shall use "%%".
For example the pattern "foo-%*.jpeg" will match all the filenames prefixed by "foo-" and terminating with ".jpeg", and "foo-%?%?%?.jpeg" will match all the filenames prefixed with "foo-", followed by a sequence of three characters, and terminating with ".jpeg".
This pattern type is deprecated in favor of glob and sequence.
Default value is glob_sequence.
Examples
ffmpeg -framerate 10 -i 'img-%03d.jpeg' out.mkv
ffmpeg -framerate 10 -start_number 100 -i 'img-%03d.jpeg' out.mkv
ffmpeg -framerate 10 -pattern_type glob -i "*.png" out.mkv
The Game Music Emu library is a collection of video game music file emulators.
See <https://bitbucket.org/mpyne/game-music-emu/overview> for more information.
It accepts the following options:
ModPlug based module demuxer
See <https://github.com/Konstanty/libmodplug>
It will export one 2-channel 16-bit 44.1 kHz audio stream. Optionally, a "pal8" 16-color video stream can be exported with or without printed metadata.
It accepts the following options:
libopenmpt based module demuxer
See <https://lib.openmpt.org/libopenmpt/> for more information.
Some files have multiple subsongs (tracks) this can be set with the subsong option.
It accepts the following options:
The default value is to let libopenmpt choose.
Demuxer for Quicktime File Format & ISO/IEC Base Media File Format (ISO/IEC 14496-12 or MPEG-4 Part 12, ISO/IEC 15444-12 or JPEG 2000 Part 12).
Registered extensions: mov, mp4, m4a, 3gp, 3g2, mj2, psp, m4b, ism, ismv, isma, f4v
Options
This demuxer accepts the following options:
Following options are available:
Unit is the track time scale. Range is 0 to UINT_MAX. Default is "UINT_MAX - 48000*10" which allows upto a 10 second dts correction for 48 kHz audio streams while accommodating 99.9% of "uint32" range.
Audible AAX
Audible AAX files are encrypted M4B files, and they can be decrypted by specifying a 4 byte activation secret.
ffmpeg -activation_bytes 1CEB00DA -i test.aax -vn -c:a copy output.mp4
MPEG-2 transport stream demuxer.
This demuxer accepts the following options:
MJPEG encapsulated in multi-part MIME demuxer.
This demuxer allows reading of MJPEG, where each frame is represented as a part of multipart/x-mixed-replace stream.
Raw video demuxer.
This demuxer allows one to read raw video data. Since there is no header specifying the assumed video parameters, the user must specify them in order to be able to decode the data correctly.
This demuxer accepts the following options:
For example to read a rawvideo file input.raw with ffplay, assuming a pixel format of "rgb24", a video size of "320x240", and a frame rate of 10 images per second, use the command:
ffplay -f rawvideo -pixel_format rgb24 -video_size 320x240 -framerate 10 input.raw
SBaGen script demuxer.
This demuxer reads the script language used by SBaGen <http://uazu.net/sbagen/> to generate binaural beats sessions. A SBG script looks like that:
-SE
a: 300-2.5/3 440+4.5/0
b: 300-2.5/0 440+4.5/3
off: -
NOW == a
+0:07:00 == b
+0:14:00 == a
+0:21:00 == b
+0:30:00 off
A SBG script can mix absolute and relative timestamps. If the script uses either only absolute timestamps (including the script start time) or only relative ones, then its layout is fixed, and the conversion is straightforward. On the other hand, if the script mixes both kind of timestamps, then the NOW reference for relative timestamps will be taken from the current time of day at the time the script is read, and the script layout will be frozen according to that reference. That means that if the script is directly played, the actual times will match the absolute timestamps up to the sound controller's clock accuracy, but if the user somehow pauses the playback or seeks, all times will be shifted accordingly.
JSON captions used for <http://www.ted.com/>.
TED does not provide links to the captions, but they can be guessed from the page. The file tools/bookmarklets.html from the FFmpeg source tree contains a bookmarklet to expose them.
This demuxer accepts the following option:
Example: convert the captions to a format most players understand:
ffmpeg -i http://www.ted.com/talks/subtitles/id/1/lang/en talk1-en.srt
Vapoursynth wrapper.
Due to security concerns, Vapoursynth scripts will not be autodetected so the input format has to be forced. For ff* CLI tools, add "-f vapoursynth" before the input "-i yourscript.vpy".
This demuxer accepts the following option:
FFmpeg is able to dump metadata from media files into a simple UTF-8-encoded INI-like text file and then load it back using the metadata muxer/demuxer.
The file format is as follows:
Next a chapter section must contain chapter start and end times in form START=num, END=num, where num is a positive integer.
A ffmetadata file might look like this:
;FFMETADATA1
title=bike\\shed
;this is a comment
artist=FFmpeg troll team
[CHAPTER]
TIMEBASE=1/1000
START=0
#chapter ends at 0:01:00
END=60000
title=chapter \#1
[STREAM]
title=multi\
line
By using the ffmetadata muxer and demuxer it is possible to extract metadata from an input file to an ffmetadata file, and then transcode the file into an output file with the edited ffmetadata file.
Extracting an ffmetadata file with ffmpeg goes as follows:
ffmpeg -i INPUT -f ffmetadata FFMETADATAFILE
Reinserting edited metadata information from the FFMETADATAFILE file can be done as:
ffmpeg -i INPUT -i FFMETADATAFILE -map_metadata 1 -codec copy OUTPUT
The libavformat library provides some generic global options, which can be set on all the protocols. In addition each protocol may support so-called private options, which are specific for that component.
Options may be set by specifying -option value in the FFmpeg tools, or by setting the value explicitly in the "AVFormatContext" options or using the libavutil/opt.h API for programmatic use.
The list of supported options follows:
Protocols are configured elements in FFmpeg that enable access to resources that require specific protocols.
When you configure your FFmpeg build, all the supported protocols are enabled by default. You can list all available ones using the configure option "--list-protocols".
You can disable all the protocols using the configure option "--disable-protocols", and selectively enable a protocol using the option "--enable-protocol=PROTOCOL", or you can disable a particular protocol using the option "--disable-protocol=PROTOCOL".
The option "-protocols" of the ff* tools will display the list of supported protocols.
All protocols accept the following options:
A description of the currently available protocols follows.
Advanced Message Queueing Protocol (AMQP) version 0-9-1 is a broker based publish-subscribe communication protocol.
FFmpeg must be compiled with --enable-librabbitmq to support AMQP. A separate AMQP broker must also be run. An example open-source AMQP broker is RabbitMQ.
After starting the broker, an FFmpeg client may stream data to the broker using the command:
ffmpeg -re -i input -f mpegts amqp://[[user]:[password]@]hostname[:port][/vhost]
Where hostname and port (default is 5672) is the address of the broker. The client may also set a user/password for authentication. The default for both fields is "guest". Name of virtual host on broker can be set with vhost. The default value is "/".
Muliple subscribers may stream from the broker using the command:
ffplay amqp://[[user]:[password]@]hostname[:port][/vhost]
In RabbitMQ all data published to the broker flows through a specific exchange, and each subscribing client has an assigned queue/buffer. When a packet arrives at an exchange, it may be copied to a client's queue depending on the exchange and routing_key fields.
The following options are supported:
Asynchronous data filling wrapper for input stream.
Fill data in a background thread, to decouple I/O operation from demux thread.
async:<URL>
async:http://host/resource
async:cache:http://host/resource
Read BluRay playlist.
The accepted options are:
Examples:
Read longest playlist from BluRay mounted to /mnt/bluray:
bluray:/mnt/bluray
Read angle 2 of playlist 4 from BluRay mounted to /mnt/bluray, start from chapter 2:
-playlist 4 -angle 2 -chapter 2 bluray:/mnt/bluray
Caching wrapper for input stream.
Cache the input stream to temporary file. It brings seeking capability to live streams.
The accepted options are:
URL Syntax is
cache:<URL>
Physical concatenation protocol.
Read and seek from many resources in sequence as if they were a unique resource.
A URL accepted by this protocol has the syntax:
concat:<URL1>|<URL2>|...|<URLN>
where URL1, URL2, ..., URLN are the urls of the resource to be concatenated, each one possibly specifying a distinct protocol.
For example to read a sequence of files split1.mpeg, split2.mpeg, split3.mpeg with ffplay use the command:
ffplay concat:split1.mpeg\|split2.mpeg\|split3.mpeg
Note that you may need to escape the character "|" which is special for many shells.
Physical concatenation protocol using a line break delimited list of resources.
Read and seek from many resources in sequence as if they were a unique resource.
A URL accepted by this protocol has the syntax:
concatf:<URL>
where URL is the url containing a line break delimited list of resources to be concatenated, each one possibly specifying a distinct protocol. Special characters must be escaped with backslash or single quotes. See the "Quoting and escaping" section in the ffmpeg-utils(1) manual.
For example to read a sequence of files split1.mpeg, split2.mpeg, split3.mpeg listed in separate lines within a file split.txt with ffplay use the command:
ffplay concatf:split.txt
Where split.txt contains the lines:
split1.mpeg
split2.mpeg
split3.mpeg
AES-encrypted stream reading protocol.
The accepted options are:
Accepted URL formats:
crypto:<URL>
crypto+<URL>
Data in-line in the URI. See <http://en.wikipedia.org/wiki/Data_URI_scheme>.
For example, to convert a GIF file given inline with ffmpeg:
ffmpeg -i "data:image/gif;base64,R0lGODdhCAAIAMIEAAAAAAAA//8AAP//AP///////////////ywAAAAACAAIAAADF0gEDLojDgdGiJdJqUX02iB4E8Q9jUMkADs=" smiley.png
File descriptor access protocol.
The accepted syntax is:
fd: -fd <file_descriptor>
If fd is not specified, by default the stdout file descriptor will be used for writing, stdin for reading. Unlike the pipe protocol, fd protocol has seek support if it corresponding to a regular file. fd protocol doesn't support pass file descriptor via URL for security.
This protocol accepts the following options:
File access protocol.
Read from or write to a file.
A file URL can have the form:
file:<filename>
where filename is the path of the file to read.
An URL that does not have a protocol prefix will be assumed to be a file URL. Depending on the build, an URL that looks like a Windows path with the drive letter at the beginning will also be assumed to be a file URL (usually not the case in builds for unix-like systems).
For example to read from a file input.mpeg with ffmpeg use the command:
ffmpeg -i file:input.mpeg output.mpeg
This protocol accepts the following options:
Many demuxers handle seekable and non-seekable resources differently, overriding this might speed up opening certain files at the cost of losing some features (e.g. accurate seeking).
FTP (File Transfer Protocol).
Read from or write to remote resources using FTP protocol.
Following syntax is required.
ftp://[user[:password]@]server[:port]/path/to/remote/resource.mpeg
This protocol accepts the following options.
NOTE: Protocol can be used as output, but it is recommended to not do it, unless special care is taken (tests, customized server configuration etc.). Different FTP servers behave in different way during seek operation. ff* tools may produce incomplete content due to server limitations.
Gopher protocol.
Gophers protocol.
The Gopher protocol with TLS encapsulation.
Read Apple HTTP Live Streaming compliant segmented stream as a uniform one. The M3U8 playlists describing the segments can be remote HTTP resources or local files, accessed using the standard file protocol. The nested protocol is declared by specifying "+proto" after the hls URI scheme name, where proto is either "file" or "http".
hls+http://host/path/to/remote/resource.m3u8
hls+file://path/to/local/resource.m3u8
Using this protocol is discouraged - the hls demuxer should work just as well (if not, please report the issues) and is more complete. To use the hls demuxer instead, simply use the direct URLs to the m3u8 files.
HTTP (Hyper Text Transfer Protocol).
This protocol accepts the following options:
When used as a server option it sets the HTTP method that is going to be expected from the client(s). If the expected and the received HTTP method do not match the client will be given a Bad Request response. When unset the HTTP method is not checked for now. This will be replaced by autodetection in the future.
# Server side (sending):
ffmpeg -i somefile.ogg -c copy -listen 1 -f ogg http://<server>:<port>
# Client side (receiving):
ffmpeg -i http://<server>:<port> -c copy somefile.ogg
# Client can also be done with wget:
wget http://<server>:<port> -O somefile.ogg
# Server side (receiving):
ffmpeg -listen 1 -i http://<server>:<port> -c copy somefile.ogg
# Client side (sending):
ffmpeg -i somefile.ogg -chunked_post 0 -c copy -f ogg http://<server>:<port>
# Client can also be done with wget:
wget --post-file=somefile.ogg http://<server>:<port>
Basic authentication sends a Base64-encoded string that contains a user name and password for the client. Base64 is not a form of encryption and should be considered the same as sending the user name and password in clear text (Base64 is a reversible encoding). If a resource needs to be protected, strongly consider using an authentication scheme other than basic authentication. HTTPS/TLS should be used with basic authentication. Without these additional security enhancements, basic authentication should not be used to protect sensitive or valuable information.
HTTP Cookies
Some HTTP requests will be denied unless cookie values are passed in with the request. The cookies option allows these cookies to be specified. At the very least, each cookie must specify a value along with a path and domain. HTTP requests that match both the domain and path will automatically include the cookie value in the HTTP Cookie header field. Multiple cookies can be delimited by a newline.
The required syntax to play a stream specifying a cookie is:
ffplay -cookies "nlqptid=nltid=tsn; path=/; domain=somedomain.com;" http://somedomain.com/somestream.m3u8
Icecast protocol (stream to Icecast servers)
This protocol accepts the following options:
icecast://[<username>[:<password>]@]<server>:<port>/<mountpoint>
InterPlanetary File System (IPFS) protocol support. One can access files stored on the IPFS network through so-called gateways. These are http(s) endpoints. This protocol wraps the IPFS native protocols (ipfs:// and ipns://) to be sent to such a gateway. Users can (and should) host their own node which means this protocol will use one's local gateway to access files on the IPFS network.
This protocol accepts the following options:
One can use this protocol in 2 ways. Using IPFS:
ffplay ipfs://<hash>
Or the IPNS protocol (IPNS is mutable IPFS):
ffplay ipns://<hash>
MMS (Microsoft Media Server) protocol over TCP.
MMS (Microsoft Media Server) protocol over HTTP.
The required syntax is:
mmsh://<server>[:<port>][/<app>][/<playpath>]
MD5 output protocol.
Computes the MD5 hash of the data to be written, and on close writes this to the designated output or stdout if none is specified. It can be used to test muxers without writing an actual file.
Some examples follow.
# Write the MD5 hash of the encoded AVI file to the file output.avi.md5.
ffmpeg -i input.flv -f avi -y md5:output.avi.md5
# Write the MD5 hash of the encoded AVI file to stdout.
ffmpeg -i input.flv -f avi -y md5:
Note that some formats (typically MOV) require the output protocol to be seekable, so they will fail with the MD5 output protocol.
UNIX pipe access protocol.
Read and write from UNIX pipes.
The accepted syntax is:
pipe:[<number>]
If fd isn't specified, number is the number corresponding to the file descriptor of the pipe (e.g. 0 for stdin, 1 for stdout, 2 for stderr). If number is not specified, by default the stdout file descriptor will be used for writing, stdin for reading.
For example to read from stdin with ffmpeg:
cat test.wav | ffmpeg -i pipe:0
# ...this is the same as...
cat test.wav | ffmpeg -i pipe:
For writing to stdout with ffmpeg:
ffmpeg -i test.wav -f avi pipe:1 | cat > test.avi
# ...this is the same as...
ffmpeg -i test.wav -f avi pipe: | cat > test.avi
This protocol accepts the following options:
Note that some formats (typically MOV), require the output protocol to be seekable, so they will fail with the pipe output protocol.
Pro-MPEG Code of Practice #3 Release 2 FEC protocol.
The Pro-MPEG CoP#3 FEC is a 2D parity-check forward error correction mechanism for MPEG-2 Transport Streams sent over RTP.
This protocol must be used in conjunction with the "rtp_mpegts" muxer and the "rtp" protocol.
The required syntax is:
-f rtp_mpegts -fec prompeg=<option>=<val>... rtp://<hostname>:<port>
The destination UDP ports are "port + 2" for the column FEC stream and "port + 4" for the row FEC stream.
This protocol accepts the following options:
Example usage:
-f rtp_mpegts -fec prompeg=l=8:d=4 rtp://<hostname>:<port>
Reliable Internet Streaming Transport protocol
The accepted options are:
Real-Time Messaging Protocol.
The Real-Time Messaging Protocol (RTMP) is used for streaming multimedia content across a TCP/IP network.
The required syntax is:
rtmp://[<username>:<password>@]<server>[:<port>][/<app>][/<instance>][/<playpath>]
The accepted parameters are:
Additionally, the following parameters can be set via command line options (or in code via "AVOption"s):
Remark: Writing to the socket is currently not optimized to minimize system calls and reduces the efficiency / effect of TCP_NODELAY.
For example to read with ffplay a multimedia resource named "sample" from the application "vod" from an RTMP server "myserver":
ffplay rtmp://myserver/vod/sample
To publish to a password protected server, passing the playpath and app names separately:
ffmpeg -re -i <input> -f flv -rtmp_playpath some/long/path -rtmp_app long/app/name rtmp://username:password@myserver/
Encrypted Real-Time Messaging Protocol.
The Encrypted Real-Time Messaging Protocol (RTMPE) is used for streaming multimedia content within standard cryptographic primitives, consisting of Diffie-Hellman key exchange and HMACSHA256, generating a pair of RC4 keys.
Real-Time Messaging Protocol over a secure SSL connection.
The Real-Time Messaging Protocol (RTMPS) is used for streaming multimedia content across an encrypted connection.
Real-Time Messaging Protocol tunneled through HTTP.
The Real-Time Messaging Protocol tunneled through HTTP (RTMPT) is used for streaming multimedia content within HTTP requests to traverse firewalls.
Encrypted Real-Time Messaging Protocol tunneled through HTTP.
The Encrypted Real-Time Messaging Protocol tunneled through HTTP (RTMPTE) is used for streaming multimedia content within HTTP requests to traverse firewalls.
Real-Time Messaging Protocol tunneled through HTTPS.
The Real-Time Messaging Protocol tunneled through HTTPS (RTMPTS) is used for streaming multimedia content within HTTPS requests to traverse firewalls.
libsmbclient permits one to manipulate CIFS/SMB network resources.
Following syntax is required.
smb://[[domain:]user[:password@]]server[/share[/path[/file]]]
This protocol accepts the following options.
For more information see: <http://www.samba.org/>.
Secure File Transfer Protocol via libssh
Read from or write to remote resources using SFTP protocol.
Following syntax is required.
sftp://[user[:password]@]server[:port]/path/to/remote/resource.mpeg
This protocol accepts the following options.
Example: Play a file stored on remote server.
ffplay sftp://user:password@server_address:22/home/user/resource.mpeg
Real-Time Messaging Protocol and its variants supported through librtmp.
Requires the presence of the librtmp headers and library during configuration. You need to explicitly configure the build with "--enable-librtmp". If enabled this will replace the native RTMP protocol.
This protocol provides most client functions and a few server functions needed to support RTMP, RTMP tunneled in HTTP (RTMPT), encrypted RTMP (RTMPE), RTMP over SSL/TLS (RTMPS) and tunneled variants of these encrypted types (RTMPTE, RTMPTS).
The required syntax is:
<rtmp_proto>://<server>[:<port>][/<app>][/<playpath>] <options>
where rtmp_proto is one of the strings "rtmp", "rtmpt", "rtmpe", "rtmps", "rtmpte", "rtmpts" corresponding to each RTMP variant, and server, port, app and playpath have the same meaning as specified for the RTMP native protocol. options contains a list of space-separated options of the form key=val.
See the librtmp manual page (man 3 librtmp) for more information.
For example, to stream a file in real-time to an RTMP server using ffmpeg:
ffmpeg -re -i myfile -f flv rtmp://myserver/live/mystream
To play the same stream using ffplay:
ffplay "rtmp://myserver/live/mystream live=1"
Real-time Transport Protocol.
The required syntax for an RTP URL is: rtp://hostname[:port][?option=val...]
port specifies the RTP port to use.
The following URL options are supported:
This is a deprecated option. Instead, localrtpport should be used.
Important notes:
Real-Time Streaming Protocol.
RTSP is not technically a protocol handler in libavformat, it is a demuxer and muxer. The demuxer supports both normal RTSP (with data transferred over RTP; this is used by e.g. Apple and Microsoft) and Real-RTSP (with data transferred over RDT).
The muxer can be used to send a stream using RTSP ANNOUNCE to a server supporting it (currently Darwin Streaming Server and Mischa Spiegelmock's <https://github.com/revmischa/rtsp-server>).
The required syntax for a RTSP url is:
rtsp://<hostname>[:<port>]/<path>
Options can be set on the ffmpeg/ffplay command line, or set in code via "AVOption"s or in "avformat_open_input".
Muxer
The following options are supported.
It accepts the following values:
Default value is 0.
The following values are accepted:
Default value is 0.
Demuxer
The following options are supported.
It accepts the following values:
Multiple lower transport protocols may be specified, in that case they are tried one at a time (if the setup of one fails, the next one is tried). For the muxer, only the tcp and udp options are supported.
The following values are accepted:
Default value is none.
The following flags are accepted:
By default it accepts all media types.
When receiving data over UDP, the demuxer tries to reorder received packets (since they may arrive out of order, or packets may get lost totally). This can be disabled by setting the maximum demuxing delay to zero (via the "max_delay" field of AVFormatContext).
When watching multi-bitrate Real-RTSP streams with ffplay, the streams to display can be chosen with "-vst" n and "-ast" n for video and audio respectively, and can be switched on the fly by pressing "v" and "a".
Examples
The following examples all make use of the ffplay and ffmpeg tools.
ffplay -max_delay 500000 -rtsp_transport udp rtsp://server/video.mp4
ffplay -rtsp_transport http rtsp://server/video.mp4
ffmpeg -re -i <input> -f rtsp -muxdelay 0.1 rtsp://server/live.sdp
ffmpeg -rtsp_flags listen -i rtsp://ownaddress/live.sdp <output>
Session Announcement Protocol (RFC 2974). This is not technically a protocol handler in libavformat, it is a muxer and demuxer. It is used for signalling of RTP streams, by announcing the SDP for the streams regularly on a separate port.
Muxer
The syntax for a SAP url given to the muxer is:
sap://<destination>[:<port>][?<options>]
The RTP packets are sent to destination on port port, or to port 5004 if no port is specified. options is a "&"-separated list. The following options are supported:
Example command lines follow.
To broadcast a stream on the local subnet, for watching in VLC:
ffmpeg -re -i <input> -f sap sap://224.0.0.255?same_port=1
Similarly, for watching in ffplay:
ffmpeg -re -i <input> -f sap sap://224.0.0.255
And for watching in ffplay, over IPv6:
ffmpeg -re -i <input> -f sap sap://[ff0e::1:2:3:4]
Demuxer
The syntax for a SAP url given to the demuxer is:
sap://[<address>][:<port>]
address is the multicast address to listen for announcements on, if omitted, the default 224.2.127.254 (sap.mcast.net) is used. port is the port that is listened on, 9875 if omitted.
The demuxers listens for announcements on the given address and port. Once an announcement is received, it tries to receive that particular stream.
Example command lines follow.
To play back the first stream announced on the normal SAP multicast address:
ffplay sap://
To play back the first stream announced on one the default IPv6 SAP multicast address:
ffplay sap://[ff0e::2:7ffe]
Stream Control Transmission Protocol.
The accepted URL syntax is:
sctp://<host>:<port>[?<options>]
The protocol accepts the following options:
Haivision Secure Reliable Transport Protocol via libsrt.
The supported syntax for a SRT URL is:
srt://<hostname>:<port>[?<options>]
options contains a list of &-separated options of the form key=val.
or
<options> srt://<hostname>:<port>
options contains a list of '-key val' options.
This protocol accepts the following options.
Special value -1: Do not drop packets on the sender at all.
Receive buffer must not be greater than ffs.
The version format in hex is 0xXXYYZZ for x.y.z in human readable form.
Stream API (default, when this option is false). In this mode you may send as many data as you wish with one sending instruction, or even use dedicated functions that read directly from a file. The internal facility will take care of any speed and congestion control. When receiving, you can also receive as many data as desired, the data not extracted will be waiting for the next call. There is no boundary between data portions in the Stream mode.
Message API. In this mode your single sending instruction passes exactly one piece of data that has boundaries (a message). Contrary to Live mode, this message may span across multiple UDP packets and the only size limitation is that it shall fit as a whole in the sending buffer. The receiver shall use as large buffer as necessary to receive the message, otherwise the message will not be given up. When the message is not complete (not all packets received or there was a packet loss) it will not be given up.
live: Set options as for live transmission. In this mode, you should send by one sending instruction only so many data that fit in one UDP packet, and limited to the value defined first in payload_size (1316 is default in this mode). There is no speed control in this mode, only the bandwidth control, if configured, in order to not exceed the bandwidth with the overhead transmission (retransmitted and control packets).
file: Set options as for non-live transmission. See messageapi for further explanations
For more information see: <https://github.com/Haivision/srt>.
Secure Real-time Transport Protocol.
The accepted options are:
Supported values:
Virtually extract a segment of a file or another stream. The underlying stream must be seekable.
Accepted options:
Examples:
Extract a chapter from a DVD VOB file (start and end sectors obtained externally and multiplied by 2048):
subfile,,start,153391104,end,268142592,,:/media/dvd/VIDEO_TS/VTS_08_1.VOB
Play an AVI file directly from a TAR archive:
subfile,,start,183241728,end,366490624,,:archive.tar
Play a MPEG-TS file from start offset till end:
subfile,,start,32815239,end,0,,:video.ts
Writes the output to multiple protocols. The individual outputs are separated by |
tee:file://path/to/local/this.avi|file://path/to/local/that.avi
Transmission Control Protocol.
The required syntax for a TCP url is:
tcp://<hostname>:<port>[?<options>]
options contains a list of &-separated options of the form key=val.
The list of supported options follows.
This option is only relevant in read mode: if no data arrived in more than this time interval, raise error.
Remark: Writing to the socket is currently not optimized to minimize system calls and reduces the efficiency / effect of TCP_NODELAY.
The following example shows how to setup a listening TCP connection with ffmpeg, which is then accessed with ffplay:
ffmpeg -i <input> -f <format> tcp://<hostname>:<port>?listen
ffplay tcp://<hostname>:<port>
Transport Layer Security (TLS) / Secure Sockets Layer (SSL)
The required syntax for a TLS/SSL url is:
tls://<hostname>:<port>[?<options>]
The following parameters can be set via command line options (or in code via "AVOption"s):
This is disabled by default since it requires a CA database to be provided by the caller in many cases.
Example command lines:
To create a TLS/SSL server that serves an input stream.
ffmpeg -i <input> -f <format> tls://<hostname>:<port>?listen&cert=<server.crt>&key=<server.key>
To play back a stream from the TLS/SSL server using ffplay:
ffplay tls://<hostname>:<port>
User Datagram Protocol.
The required syntax for an UDP URL is:
udp://<hostname>:<port>[?<options>]
options contains a list of &-separated options of the form key=val.
In case threading is enabled on the system, a circular buffer is used to store the incoming data, which allows one to reduce loss of data due to UDP socket buffer overruns. The fifo_size and overrun_nonfatal options are related to this buffer.
The list of supported options follows.
This option is only relevant in read mode: if no data arrived in more than this time interval, raise error.
Note that broadcasting may not work properly on networks having a broadcast storm protection.
Examples
ffmpeg -i <input> -f <format> udp://<hostname>:<port>
ffmpeg -i <input> -f mpegts udp://<hostname>:<port>?pkt_size=188&buffer_size=65535
ffmpeg -i udp://[<multicast-address>]:<port> ...
Unix local socket
The required syntax for a Unix socket URL is:
unix://<filepath>
The following parameters can be set via command line options (or in code via "AVOption"s):
ZeroMQ asynchronous messaging using the libzmq library.
This library supports unicast streaming to multiple clients without relying on an external server.
The required syntax for streaming or connecting to a stream is:
zmq:tcp://ip-address:port
Example: Create a localhost stream on port 5555:
ffmpeg -re -i input -f mpegts zmq:tcp://127.0.0.1:5555
Multiple clients may connect to the stream using:
ffplay zmq:tcp://127.0.0.1:5555
Streaming to multiple clients is implemented using a ZeroMQ Pub-Sub pattern. The server side binds to a port and publishes data. Clients connect to the server (via IP address/port) and subscribe to the stream. The order in which the server and client start generally does not matter.
ffmpeg must be compiled with the --enable-libzmq option to support this protocol.
Options can be set on the ffmpeg/ffplay command line. The following options are supported:
The libavdevice library provides the same interface as libavformat. Namely, an input device is considered like a demuxer, and an output device like a muxer, and the interface and generic device options are the same provided by libavformat (see the ffmpeg-formats manual).
In addition each input or output device may support so-called private options, which are specific for that component.
Options may be set by specifying -option value in the FFmpeg tools, or by setting the value explicitly in the device "AVFormatContext" options or using the libavutil/opt.h API for programmatic use.
Input devices are configured elements in FFmpeg which enable accessing the data coming from a multimedia device attached to your system.
When you configure your FFmpeg build, all the supported input devices are enabled by default. You can list all available ones using the configure option "--list-indevs".
You can disable all the input devices using the configure option "--disable-indevs", and selectively enable an input device using the option "--enable-indev=INDEV", or you can disable a particular input device using the option "--disable-indev=INDEV".
The option "-devices" of the ff* tools will display the list of supported input devices.
A description of the currently available input devices follows.
ALSA (Advanced Linux Sound Architecture) input device.
To enable this input device during configuration you need libasound installed on your system.
This device allows capturing from an ALSA device. The name of the device to capture has to be an ALSA card identifier.
An ALSA identifier has the syntax:
hw:<CARD>[,<DEV>[,<SUBDEV>]]
where the DEV and SUBDEV components are optional.
The three arguments (in order: CARD,DEV,SUBDEV) specify card number or identifier, device number and subdevice number (-1 means any).
To see the list of cards currently recognized by your system check the files /proc/asound/cards and /proc/asound/devices.
For example to capture with ffmpeg from an ALSA device with card id 0, you may run the command:
ffmpeg -f alsa -i hw:0 alsaout.wav
For more information see: <http://www.alsa-project.org/alsa-doc/alsa-lib/pcm.html>
Options
Android camera input device.
This input devices uses the Android Camera2 NDK API which is available on devices with API level 24+. The availability of android_camera is autodetected during configuration.
This device allows capturing from all cameras on an Android device, which are integrated into the Camera2 NDK API.
The available cameras are enumerated internally and can be selected with the camera_index parameter. The input file string is discarded.
Generally the back facing camera has index 0 while the front facing camera has index 1.
Options
AVFoundation input device.
AVFoundation is the currently recommended framework by Apple for streamgrabbing on OSX >= 10.7 as well as on iOS.
The input filename has to be given in the following syntax:
-i "[[VIDEO]:[AUDIO]]"
The first entry selects the video input while the latter selects the audio input. The stream has to be specified by the device name or the device index as shown by the device list. Alternatively, the video and/or audio input device can be chosen by index using the
B<-video_device_index E<lt>INDEXE<gt>>
and/or
B<-audio_device_index E<lt>INDEXE<gt>>
, overriding any device name or index given in the input filename.
All available devices can be enumerated by using -list_devices true, listing all device names and corresponding indices.
There are two device name aliases:
Options
AVFoundation supports the following options:
Examples
$ ffmpeg -f avfoundation -list_devices true -i ""
$ ffmpeg -f avfoundation -i "0:0" out.avi
$ ffmpeg -f avfoundation -video_device_index 2 -i ":1" out.avi
$ ffmpeg -f avfoundation -pixel_format bgr0 -i "default:none" out.avi
$ ffmpeg -f avfoundation -capture_raw_data true -i "zr100:none" out.dv
BSD video input device.
Options
The decklink input device provides capture capabilities for Blackmagic DeckLink devices.
To enable this input device, you need the Blackmagic DeckLink SDK and you need to configure with the appropriate "--extra-cflags" and "--extra-ldflags". On Windows, you need to run the IDL files through widl.
DeckLink is very picky about the formats it supports. Pixel format of the input can be set with raw_format. Framerate and video size must be determined for your device with -list_formats 1. Audio sample rate is always 48 kHz and the number of channels can be 2, 8 or 16. Note that all audio channels are bundled in one single audio track.
Options
This option is a bitmask of the SD PAL VBI lines captured, specifically lines 6 to 22, and lines 318 to 335. Line 6 is the LSB in the mask. Selected lines which do not contain teletext information will be ignored. You can use the special all constant to select all possible lines, or standard to skip lines 6, 318 and 319, which are not compatible with all receivers.
For SD sources, ffmpeg needs to be compiled with "--enable-libzvbi". For HD sources, on older (pre-4K) DeckLink card models you have to capture in 10 bit mode.
Note: DeckLink SDK 11.0 have replaced the duplex property by a profile property. For the DeckLink Duo 2 and DeckLink Quad 2, a profile is shared between any 2 sub-devices that utilize the same connectors. For the DeckLink 8K Pro, a profile is shared between all 4 sub-devices. So DeckLink 8K Pro support four profiles.
Valid profile modes for DeckLink 8K Pro(with DeckLink SDK >= 11.0): one_sub_device_full, one_sub_device_half, two_sub_device_full, four_sub_device_half
Valid profile modes for DeckLink Quad 2 and DeckLink Duo 2: half, full
In order to properly support 50/60 fps timecodes, the ordering of the queried timecode types for rp188any is HFR, VITC1, VITC2 and LTC for >30 fps content. Note that this is slightly different to the ordering used by the DeckLink API, which is HFR, VITC1, LTC, VITC2.
Examples
ffmpeg -sources decklink
ffmpeg -f decklink -list_formats 1 -i 'Intensity Pro'
ffmpeg -format_code Hi50 -f decklink -i 'Intensity Pro' -c:a copy -c:v copy output.avi
ffmpeg -raw_format yuv422p10 -format_code Hi50 -f decklink -i 'UltraStudio Mini Recorder' -c:a copy -c:v copy output.avi
ffmpeg -channels 16 -format_code Hi50 -f decklink -i 'UltraStudio Mini Recorder' -c:a copy -c:v copy output.avi
Windows DirectShow input device.
DirectShow support is enabled when FFmpeg is built with the mingw-w64 project. Currently only audio and video devices are supported.
Multiple devices may be opened as separate inputs, but they may also be opened on the same input, which should improve synchronism between them.
The input name should be in the format:
<TYPE>=<NAME>[:<TYPE>=<NAME>]
where TYPE can be either audio or video, and NAME is the device's name or alternative name..
Options
If no options are specified, the device's defaults are used. If the device does not support the requested options, it will fail to open.
Examples
$ ffmpeg -list_devices true -f dshow -i dummy
$ ffmpeg -f dshow -i video="Camera"
$ ffmpeg -f dshow -video_device_number 1 -i video="Camera"
$ ffmpeg -f dshow -i video="Camera":audio="Microphone"
$ ffmpeg -list_options true -f dshow -i video="Camera"
$ ffmpeg -f dshow -audio_pin_name "Audio Out" -video_pin_name 2 -i video=video="@device_pnp_\\?\pci#ven_1a0a&dev_6200&subsys_62021461&rev_01#4&e2c7dd6&0&00e1#{65e8773d-8f56-11d0-a3b9-00a0c9223196}\{ca465100-deb0-4d59-818f-8c477184adf6}":audio="Microphone"
$ ffmpeg -f dshow -show_video_device_dialog true -crossbar_video_input_pin_number 0
-crossbar_audio_input_pin_number 3 -i video="AVerMedia BDA Analog Capture":audio="AVerMedia BDA Analog Capture"
Linux framebuffer input device.
The Linux framebuffer is a graphic hardware-independent abstraction layer to show graphics on a computer monitor, typically on the console. It is accessed through a file device node, usually /dev/fb0.
For more detailed information read the file Documentation/fb/framebuffer.txt included in the Linux source tree.
See also <http://linux-fbdev.sourceforge.net/>, and fbset(1).
To record from the framebuffer device /dev/fb0 with ffmpeg:
ffmpeg -f fbdev -framerate 10 -i /dev/fb0 out.avi
You can take a single screenshot image with the command:
ffmpeg -f fbdev -framerate 1 -i /dev/fb0 -frames:v 1 screenshot.jpeg
Options
Win32 GDI-based screen capture device.
This device allows you to capture a region of the display on Windows.
There are two options for the input filename:
desktop
or
title=<window_title>
The first option will capture the entire desktop, or a fixed region of the desktop. The second option will instead capture the contents of a single window, regardless of its position on the screen.
For example, to grab the entire desktop using ffmpeg:
ffmpeg -f gdigrab -framerate 6 -i desktop out.mpg
Grab a 640x480 region at position "10,20":
ffmpeg -f gdigrab -framerate 6 -offset_x 10 -offset_y 20 -video_size vga -i desktop out.mpg
Grab the contents of the window named "Calculator"
ffmpeg -f gdigrab -framerate 6 -i title=Calculator out.mpg
Options
If show_region is specified with 1, then the grabbing region will be indicated on screen. With this option, it is easy to know what is being grabbed if only a portion of the screen is grabbed.
Note that show_region is incompatible with grabbing the contents of a single window.
For example:
ffmpeg -f gdigrab -show_region 1 -framerate 6 -video_size cif -offset_x 10 -offset_y 20 -i desktop out.mpg
Note that the offset calculation is from the top left corner of the primary monitor on Windows. If you have a monitor positioned to the left of your primary monitor, you will need to use a negative offset_x value to move the region to that monitor.
Note that the offset calculation is from the top left corner of the primary monitor on Windows. If you have a monitor positioned above your primary monitor, you will need to use a negative offset_y value to move the region to that monitor.
FireWire DV/HDV input device using libiec61883.
To enable this input device, you need libiec61883, libraw1394 and libavc1394 installed on your system. Use the configure option "--enable-libiec61883" to compile with the device enabled.
The iec61883 capture device supports capturing from a video device connected via IEEE1394 (FireWire), using libiec61883 and the new Linux FireWire stack (juju). This is the default DV/HDV input method in Linux Kernel 2.6.37 and later, since the old FireWire stack was removed.
Specify the FireWire port to be used as input file, or "auto" to choose the first port connected.
Options
Examples
ffplay -f iec61883 -i auto
ffmpeg -f iec61883 -i auto -dvbuffer 100000 out.mpg
JACK input device.
To enable this input device during configuration you need libjack installed on your system.
A JACK input device creates one or more JACK writable clients, one for each audio channel, with name client_name:input_N, where client_name is the name provided by the application, and N is a number which identifies the channel. Each writable client will send the acquired data to the FFmpeg input device.
Once you have created one or more JACK readable clients, you need to connect them to one or more JACK writable clients.
To connect or disconnect JACK clients you can use the jack_connect and jack_disconnect programs, or do it through a graphical interface, for example with qjackctl.
To list the JACK clients and their properties you can invoke the command jack_lsp.
Follows an example which shows how to capture a JACK readable client with ffmpeg.
# Create a JACK writable client with name "ffmpeg".
$ ffmpeg -f jack -i ffmpeg -y out.wav
# Start the sample jack_metro readable client.
$ jack_metro -b 120 -d 0.2 -f 4000
# List the current JACK clients.
$ jack_lsp -c
system:capture_1
system:capture_2
system:playback_1
system:playback_2
ffmpeg:input_1
metro:120_bpm
# Connect metro to the ffmpeg writable client.
$ jack_connect metro:120_bpm ffmpeg:input_1
For more information read: <http://jackaudio.org/>
Options
KMS video input device.
Captures the KMS scanout framebuffer associated with a specified CRTC or plane as a DRM object that can be passed to other hardware functions.
Requires either DRM master or CAP_SYS_ADMIN to run.
If you don't understand what all of that means, you probably don't want this. Look at x11grab instead.
Options
Examples
ffmpeg -f kmsgrab -i - -vf 'hwdownload,format=bgr0' output.mp4
ffmpeg -crtc_id 42 -framerate 60 -f kmsgrab -i - -vf 'hwmap=derive_device=vaapi,scale_vaapi=w=1920:h=1080:format=nv12' -c:v h264_vaapi output.mp4
ffmpeg -f kmsgrab -i - -vf 'hwmap=derive_device=vaapi,crop=960:540:480:270,scale_vaapi=960:540:nv12' -c:v h264_vaapi output.mp4
Libavfilter input virtual device.
This input device reads data from the open output pads of a libavfilter filtergraph.
For each filtergraph open output, the input device will create a corresponding stream which is mapped to the generated output. The filtergraph is specified through the option graph.
Options
The suffix "+subcc" can be appended to the output label to create an extra stream with the closed captions packets attached to that output (experimental; only for EIA-608 / CEA-708 for now). The subcc streams are created after all the normal streams, in the order of the corresponding stream. For example, if there is "out19+subcc", "out7+subcc" and up to "out42", the stream #43 is subcc for stream #7 and stream #44 is subcc for stream #19.
If not specified defaults to the filename specified for the input device.
Examples
ffplay -f lavfi -graph "color=c=pink [out0]" dummy
ffplay -f lavfi color=c=pink
ffplay -f lavfi -graph "testsrc [out0]; testsrc,hflip [out1]; testsrc,negate [out2]" test3
ffplay -f lavfi "amovie=test.wav"
ffplay -f lavfi "movie=test.avi[out0];amovie=test.wav[out1]"
ffmpeg -f lavfi -i "movie=test.ts[out0+subcc]" -map v frame%08d.png -map s -c copy -f rawvideo subcc.bin
Audio-CD input device based on libcdio.
To enable this input device during configuration you need libcdio installed on your system. It requires the configure option "--enable-libcdio".
This device allows playing and grabbing from an Audio-CD.
For example to copy with ffmpeg the entire Audio-CD in /dev/sr0, you may run the command:
ffmpeg -f libcdio -i /dev/sr0 cd.wav
Options
The speed is specified CD-ROM speed units. The speed is set through the libcdio "cdio_cddap_speed_set" function. On many CD-ROM drives, specifying a value too large will result in using the fastest speed.
Default value is disable.
For more information about the available recovery modes, consult the paranoia project documentation.
IIDC1394 input device, based on libdc1394 and libraw1394.
Requires the configure option "--enable-libdc1394".
Options
The OpenAL input device provides audio capture on all systems with a working OpenAL 1.1 implementation.
To enable this input device during configuration, you need OpenAL headers and libraries installed on your system, and need to configure FFmpeg with "--enable-openal".
OpenAL headers and libraries should be provided as part of your OpenAL implementation, or as an additional download (an SDK). Depending on your installation you may need to specify additional flags via the "--extra-cflags" and "--extra-ldflags" for allowing the build system to locate the OpenAL headers and libraries.
An incomplete list of OpenAL implementations follows:
This device allows one to capture from an audio input device handled through OpenAL.
You need to specify the name of the device to capture in the provided filename. If the empty string is provided, the device will automatically select the default device. You can get the list of the supported devices by using the option list_devices.
Options
Examples
Print the list of OpenAL supported devices and exit:
$ ffmpeg -list_devices true -f openal -i dummy out.ogg
Capture from the OpenAL device DR-BT101 via PulseAudio:
$ ffmpeg -f openal -i 'DR-BT101 via PulseAudio' out.ogg
Capture from the default device (note the empty string '' as filename):
$ ffmpeg -f openal -i '' out.ogg
Capture from two devices simultaneously, writing to two different files, within the same ffmpeg command:
$ ffmpeg -f openal -i 'DR-BT101 via PulseAudio' out1.ogg -f openal -i 'ALSA Default' out2.ogg
Note: not all OpenAL implementations support multiple simultaneous capture - try the latest OpenAL Soft if the above does not work.
Open Sound System input device.
The filename to provide to the input device is the device node representing the OSS input device, and is usually set to /dev/dsp.
For example to grab from /dev/dsp using ffmpeg use the command:
ffmpeg -f oss -i /dev/dsp /tmp/oss.wav
For more information about OSS see: <http://manuals.opensound.com/usersguide/dsp.html>
Options
PulseAudio input device.
To enable this output device you need to configure FFmpeg with "--enable-libpulse".
The filename to provide to the input device is a source device or the string "default"
To list the PulseAudio source devices and their properties you can invoke the command pactl list sources.
More information about PulseAudio can be found on <http://www.pulseaudio.org>.
Options
Examples
Record a stream from default device:
ffmpeg -f pulse -i default /tmp/pulse.wav
sndio input device.
To enable this input device during configuration you need libsndio installed on your system.
The filename to provide to the input device is the device node representing the sndio input device, and is usually set to /dev/audio0.
For example to grab from /dev/audio0 using ffmpeg use the command:
ffmpeg -f sndio -i /dev/audio0 /tmp/oss.wav
Options
Video4Linux2 input video device.
"v4l2" can be used as alias for "video4linux2".
If FFmpeg is built with v4l-utils support (by using the "--enable-libv4l2" configure option), it is possible to use it with the "-use_libv4l2" input device option.
The name of the device to grab is a file device node, usually Linux systems tend to automatically create such nodes when the device (e.g. an USB webcam) is plugged into the system, and has a name of the kind /dev/videoN, where N is a number associated to the device.
Video4Linux2 devices usually support a limited set of widthxheight sizes and frame rates. You can check which are supported using -list_formats all for Video4Linux2 devices. Some devices, like TV cards, support one or more standards. It is possible to list all the supported standards using -list_standards all.
The time base for the timestamps is 1 microsecond. Depending on the kernel version and configuration, the timestamps may be derived from the real time clock (origin at the Unix Epoch) or the monotonic clock (origin usually at boot time, unaffected by NTP or manual changes to the clock). The -timestamps abs or -ts abs option can be used to force conversion into the real time clock.
Some usage examples of the video4linux2 device with ffmpeg and ffplay:
ffplay -f video4linux2 -list_formats all /dev/video0
ffplay -f video4linux2 -framerate 30 -video_size hd720 /dev/video0
ffmpeg -f video4linux2 -input_format mjpeg -i /dev/video0 out.mpeg
For more information about Video4Linux, check <http://linuxtv.org/>.
Options
Available values are:
Available values are:
Available values are:
Default value is "default".
VfW (Video for Windows) capture input device.
The filename passed as input is the capture driver number, ranging from 0 to 9. You may use "list" as filename to print a list of drivers. Any other filename will be interpreted as device number 0.
Options
X11 video input device.
To enable this input device during configuration you need libxcb installed on your system. It will be automatically detected during configuration.
This device allows one to capture a region of an X11 display.
The filename passed as input has the syntax:
[<hostname>]:<display_number>.<screen_number>[+<x_offset>,<y_offset>]
hostname:display_number.screen_number specifies the X11 display name of the screen to grab from. hostname can be omitted, and defaults to "localhost". The environment variable DISPLAY contains the default display name.
x_offset and y_offset specify the offsets of the grabbed area with respect to the top-left border of the X11 screen. They default to 0.
Check the X11 documentation (e.g. man X) for more detailed information.
Use the xdpyinfo program for getting basic information about the properties of your X11 display (e.g. grep for "name" or "dimensions").
For example to grab from :0.0 using ffmpeg:
ffmpeg -f x11grab -framerate 25 -video_size cif -i :0.0 out.mpg
Grab at position "10,20":
ffmpeg -f x11grab -framerate 25 -video_size cif -i :0.0+10,20 out.mpg
Options
When it is specified with "centered", the grabbing region follows the mouse pointer and keeps the pointer at the center of region; otherwise, the region follows only when the mouse pointer reaches within PIXELS (greater than zero) to the edge of region.
For example:
ffmpeg -f x11grab -follow_mouse centered -framerate 25 -video_size cif -i :0.0 out.mpg
To follow only when the mouse pointer reaches within 100 pixels to edge:
ffmpeg -f x11grab -follow_mouse 100 -framerate 25 -video_size cif -i :0.0 out.mpg
If show_region is specified with 1, then the grabbing region will be indicated on screen. With this option, it is easy to know what is being grabbed if only a portion of the screen is grabbed.
For example:
ffmpeg -f x11grab -show_region 1 -framerate 25 -video_size cif -i :0.0+10,20 out.mpg
With follow_mouse:
ffmpeg -f x11grab -follow_mouse centered -show_region 1 -framerate 25 -video_size cif -i :0.0 out.mpg
The id of a window can be found using the xwininfo program, possibly with options -tree and -root.
If the window is later enlarged, the new area is not recorded. Video ends when the window is closed, unmapped (i.e., iconified) or shrunk beyond the video size (which defaults to the initial window size).
This option disables options follow_mouse and select_region.
The audio resampler supports the following named options.
Options may be set by specifying -option value in the FFmpeg tools, option=value for the aresample filter, by setting the value explicitly in the "SwrContext" options or using the libavutil/opt.h API for programmatic use.
See the Channel Layout section in the ffmpeg-utils(1) manual for the required syntax.
It supports the following individual flags:
Supported values:
Supported values:
It accepts the following values:
Default value is "none".
It accepts the following values:
The video scaler supports the following named options.
Options may be set by specifying -option value in the FFmpeg tools, with a few API-only exceptions noted below. For programmatic use, they can be set explicitly in the "SwsContext" options or through the libavutil/opt.h API.
It accepts the following values:
Filtering in FFmpeg is enabled through the libavfilter library.
In libavfilter, a filter can have multiple inputs and multiple outputs. To illustrate the sorts of things that are possible, we consider the following filtergraph.
[main]
input --> split ---------------------> overlay --> output
| ^
|[tmp] [flip]|
+-----> crop --> vflip -------+
This filtergraph splits the input stream in two streams, then sends one stream through the crop filter and the vflip filter, before merging it back with the other stream by overlaying it on top. You can use the following command to achieve this:
ffmpeg -i INPUT -vf "split [main][tmp]; [tmp] crop=iw:ih/2:0:0, vflip [flip]; [main][flip] overlay=0:H/2" OUTPUT
The result will be that the top half of the video is mirrored onto the bottom half of the output video.
Filters in the same linear chain are separated by commas, and distinct linear chains of filters are separated by semicolons. In our example, crop,vflip are in one linear chain, split and overlay are separately in another. The points where the linear chains join are labelled by names enclosed in square brackets. In the example, the split filter generates two outputs that are associated to the labels [main] and [tmp].
The stream sent to the second output of split, labelled as [tmp], is processed through the crop filter, which crops away the lower half part of the video, and then vertically flipped. The overlay filter takes in input the first unchanged output of the split filter (which was labelled as [main]), and overlay on its lower half the output generated by the crop,vflip filterchain.
Some filters take in input a list of parameters: they are specified after the filter name and an equal sign, and are separated from each other by a colon.
There exist so-called source filters that do not have an audio/video input, and sink filters that will not have audio/video output.
The graph2dot program included in the FFmpeg tools directory can be used to parse a filtergraph description and issue a corresponding textual representation in the dot language.
Invoke the command:
graph2dot -h
to see how to use graph2dot.
You can then pass the dot description to the dot program (from the graphviz suite of programs) and obtain a graphical representation of the filtergraph.
For example the sequence of commands:
echo <GRAPH_DESCRIPTION> | \
tools/graph2dot -o graph.tmp && \
dot -Tpng graph.tmp -o graph.png && \
display graph.png
can be used to create and display an image representing the graph described by the GRAPH_DESCRIPTION string. Note that this string must be a complete self-contained graph, with its inputs and outputs explicitly defined. For example if your command line is of the form:
ffmpeg -i infile -vf scale=640:360 outfile
your GRAPH_DESCRIPTION string will need to be of the form:
nullsrc,scale=640:360,nullsink
you may also need to set the nullsrc parameters and add a format filter in order to simulate a specific input file.
A filtergraph is a directed graph of connected filters. It can contain cycles, and there can be multiple links between a pair of filters. Each link has one input pad on one side connecting it to one filter from which it takes its input, and one output pad on the other side connecting it to one filter accepting its output.
Each filter in a filtergraph is an instance of a filter class registered in the application, which defines the features and the number of input and output pads of the filter.
A filter with no input pads is called a "source", and a filter with no output pads is called a "sink".
A filtergraph has a textual representation, which is recognized by the -filter/-vf/-af and -filter_complex options in ffmpeg and -vf/-af in ffplay, and by the avfilter_graph_parse_ptr() function defined in libavfilter/avfilter.h.
A filterchain consists of a sequence of connected filters, each one connected to the previous one in the sequence. A filterchain is represented by a list of ","-separated filter descriptions.
A filtergraph consists of a sequence of filterchains. A sequence of filterchains is represented by a list of ";"-separated filterchain descriptions.
A filter is represented by a string of the form: [in_link_1]...[in_link_N]filter_name@id=arguments[out_link_1]...[out_link_M]
filter_name is the name of the filter class of which the described filter is an instance of, and has to be the name of one of the filter classes registered in the program optionally followed by "@id". The name of the filter class is optionally followed by a string "=arguments".
arguments is a string which contains the parameters used to initialize the filter instance. It may have one of two forms:
If the option value itself is a list of items (e.g. the "format" filter takes a list of pixel formats), the items in the list are usually separated by |.
The list of arguments can be quoted using the character ' as initial and ending mark, and the character \ for escaping the characters within the quoted text; otherwise the argument string is considered terminated when the next special character (belonging to the set []=;,) is encountered.
A special syntax implemented in the ffmpeg CLI tool allows loading option values from files. This is done be prepending a slash '/' to the option name, then the supplied value is interpreted as a path from which the actual value is loaded. E.g.
ffmpeg -i <INPUT> -vf drawtext=/text=/tmp/some_text <OUTPUT>
will load the text to be drawn from /tmp/some_text. API users wishing to implement a similar feature should use the "avfilter_graph_segment_*()" functions together with custom IO code.
The name and arguments of the filter are optionally preceded and followed by a list of link labels. A link label allows one to name a link and associate it to a filter output or input pad. The preceding labels in_link_1 ... in_link_N, are associated to the filter input pads, the following labels out_link_1 ... out_link_M, are associated to the output pads.
When two link labels with the same name are found in the filtergraph, a link between the corresponding input and output pad is created.
If an output pad is not labelled, it is linked by default to the first unlabelled input pad of the next filter in the filterchain. For example in the filterchain
nullsrc, split[L1], [L2]overlay, nullsink
the split filter instance has two output pads, and the overlay filter instance two input pads. The first output pad of split is labelled "L1", the first input pad of overlay is labelled "L2", and the second output pad of split is linked to the second input pad of overlay, which are both unlabelled.
In a filter description, if the input label of the first filter is not specified, "in" is assumed; if the output label of the last filter is not specified, "out" is assumed.
In a complete filterchain all the unlabelled filter input and output pads must be connected. A filtergraph is considered valid if all the filter input and output pads of all the filterchains are connected.
Leading and trailing whitespaces (space, tabs, or line feeds) separating tokens in the filtergraph specification are ignored. This means that the filtergraph can be expressed using empty lines and spaces to improve redability.
For example, the filtergraph:
testsrc,split[L1],hflip[L2];[L1][L2] hstack
can be represented as:
testsrc,
split [L1], hflip [L2];
[L1][L2] hstack
Libavfilter will automatically insert scale filters where format conversion is required. It is possible to specify swscale flags for those automatically inserted scalers by prepending "sws_flags=flags;" to the filtergraph description.
Here is a BNF description of the filtergraph syntax:
<NAME> ::= sequence of alphanumeric characters and '_'
<FILTER_NAME> ::= <NAME>["@"<NAME>]
<LINKLABEL> ::= "[" <NAME> "]"
<LINKLABELS> ::= <LINKLABEL> [<LINKLABELS>]
<FILTER_ARGUMENTS> ::= sequence of chars (possibly quoted)
<FILTER> ::= [<LINKLABELS>] <FILTER_NAME> ["=" <FILTER_ARGUMENTS>] [<LINKLABELS>]
<FILTERCHAIN> ::= <FILTER> [,<FILTERCHAIN>]
<FILTERGRAPH> ::= [sws_flags=<flags>;] <FILTERCHAIN> [;<FILTERGRAPH>]
Filtergraph description composition entails several levels of escaping. See the "Quoting and escaping" section in the ffmpeg-utils(1) manual for more information about the employed escaping procedure.
A first level escaping affects the content of each filter option value, which may contain the special character ":" used to separate values, or one of the escaping characters "\'".
A second level escaping affects the whole filter description, which may contain the escaping characters "\'" or the special characters "[],;" used by the filtergraph description.
Finally, when you specify a filtergraph on a shell commandline, you need to perform a third level escaping for the shell special characters contained within it.
For example, consider the following string to be embedded in the drawtext filter description text value:
this is a 'string': may contain one, or more, special characters
This string contains the "'" special escaping character, and the ":" special character, so it needs to be escaped in this way:
text=this is a \'string\'\: may contain one, or more, special characters
A second level of escaping is required when embedding the filter description in a filtergraph description, in order to escape all the filtergraph special characters. Thus the example above becomes:
drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
(note that in addition to the "\'" escaping special characters, also "," needs to be escaped).
Finally an additional level of escaping is needed when writing the filtergraph description in a shell command, which depends on the escaping rules of the adopted shell. For example, assuming that "\" is special and needs to be escaped with another "\", the previous string will finally result in:
-vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
In order to avoid cumbersome escaping when using a commandline tool accepting a filter specification as input, it is advisable to avoid direct inclusion of the filter or options specification in the shell.
For example, in case of the drawtext filter, you might prefer to use the textfile option in place of text to specify the text to render.
When using the ffmpeg tool, you might consider to use the -filter_script option or -filter_complex_script option.
Some filters support a generic enable option. For the filters supporting timeline editing, this option can be set to an expression which is evaluated before sending a frame to the filter. If the evaluation is non-zero, the filter will be enabled, otherwise the frame will be sent unchanged to the next filter in the filtergraph.
The expression accepts the following values:
Additionally, these filters support an enable command that can be used to re-define the expression.
Like any other filtering option, the enable option follows the same rules.
For example, to enable a blur filter (smartblur) from 10 seconds to 3 minutes, and a curves filter starting at 3 seconds:
smartblur = enable='between(t,10,3*60)',
curves = enable='gte(t,3)' : preset=cross_process
See "ffmpeg -filters" to view which filters have timeline support.
Some options can be changed during the operation of the filter using a command. These options are marked 'T' on the output of ffmpeg -h filter=<name of filter>. The name of the command is the name of the option and the argument is the new value.
Some filters with several inputs support a common set of options. These options can only be set by name, not with the short notation.
When you configure your FFmpeg build, you can disable any of the existing filters using "--disable-filters". The configure output will show the audio filters included in your build.
Below is a description of the currently available audio filters.
A compressor is mainly used to reduce the dynamic range of a signal. Especially modern music is mostly compressed at a high ratio to improve the overall loudness. It's done to get the highest attention of a listener, "fatten" the sound and bring more "power" to the track. If a signal is compressed too much it may sound dull or "dead" afterwards or it may start to "pump" (which could be a powerful effect but can also destroy a track completely). The right compression is the key to reach a professional sound and is the high art of mixing and mastering. Because of its complex settings it may take a long time to get the right feeling for this kind of effect.
Compression is done by detecting the volume above a chosen level "threshold" and dividing it by the factor set with "ratio". So if you set the threshold to -12dB and your signal reaches -6dB a ratio of 2:1 will result in a signal at -9dB. Because an exact manipulation of the signal would cause distortion of the waveform the reduction can be levelled over the time. This is done by setting "Attack" and "Release". "attack" determines how long the signal has to rise above the threshold before any reduction will occur and "release" sets the time the signal has to fall below the threshold to reduce the reduction again. Shorter signals than the chosen attack time will be left untouched. The overall reduction of the signal can be made up afterwards with the "makeup" setting. So compressing the peaks of a signal about 6dB and raising the makeup to this level results in a signal twice as loud than the source. To gain a softer entry in the compression the "knee" flattens the hard edge at the threshold in the range of the chosen decibels.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Simple audio dynamic range compression/expansion filter.
The filter accepts the following options:
Copy the input audio source unchanged to the output. This is mainly useful for testing purposes.
Apply cross fade from one input audio stream to another input audio stream. The cross fade is applied for specified duration near the end of first stream.
The filter accepts the following options:
For description of available curve types see afade filter description.
Examples
ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
Split audio stream into several bands.
This filter splits audio stream into two or more frequency ranges. Summing all streams back will give flat output.
The filter accepts the following options:
Default is 4th.
Default value is "auto".
Examples
ffmpeg -i in.flac -filter_complex 'acrossover=split=1500[LOW][HIGH]' -map '[LOW]' low.wav -map '[HIGH]' high.wav
ffmpeg -i in.flac -filter_complex 'acrossover=split=1500:order=8th[LOW][HIGH]' -map '[LOW]' low.wav -map '[HIGH]' high.wav
ffmpeg -i in.flac -filter_complex 'acrossover=split=1500 8000:order=8th[LOW][MID][HIGH]' -map '[LOW]' low.wav -map '[MID]' mid.wav -map '[HIGH]' high.wav
Reduce audio bit resolution.
This filter is bit crusher with enhanced functionality. A bit crusher is used to audibly reduce number of bits an audio signal is sampled with. This doesn't change the bit depth at all, it just produces the effect. Material reduced in bit depth sounds more harsh and "digital". This filter is able to even round to continuous values instead of discrete bit depths. Additionally it has a D/C offset which results in different crushing of the lower and the upper half of the signal. An Anti-Aliasing setting is able to produce "softer" crushing sounds.
Another feature of this filter is the logarithmic mode. This setting switches from linear distances between bits to logarithmic ones. The result is a much more "natural" sounding crusher which doesn't gate low signals for example. The human ear has a logarithmic perception, so this kind of crushing is much more pleasant. Logarithmic crushing is also able to get anti-aliased.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Delay audio filtering until a given wallclock timestamp. See the cue filter.
Remove impulsive noise from input audio.
Samples detected as impulsive noise are replaced by interpolated samples using autoregressive modelling.
It accepts the following values:
Default value is "a".
Remove clipped samples from input audio.
Samples detected as clipped are replaced by interpolated samples using autoregressive modelling.
It accepts the following values:
Default value is "a".
Apply decorrelation to input audio stream.
The filter accepts the following options:
Delay one or more audio channels.
Samples in delayed channel are filled with silence.
The filter accepts the following option:
Examples
adelay=1500|0|500
adelay=0|500S|700S
adelay=delays=64S:all=1
Remedy denormals in audio by adding extremely low-level noise.
This filter shall be placed before any filter that can produce denormals.
A description of the accepted parameters follows.
Default is "dc".
Commands
This filter supports the all above options as commands.
Compute derivative/integral of audio stream.
Applying both filters one after another produces original audio.
Apply spectral dynamic range controller filter to input audio stream.
A description of the accepted options follows.
The expression can contain the following constants:
Default value is "p".
Commands
This filter supports the all above options as commands.
Examples
adrc=transfer='if(gt(p,-50),-50+(p-(-50))/6,p)':attack=50:release=100
adrc=transfer='if(gt(p,-50),-50+(p-(-50))/2,p)':attack=50:release=100:channels=FC
adrc=transfer='if(lte(p,-85),p-800,p)':attack=1:release=5
adrc=transfer='if(lt(p,-10),-10+(p-(-10))*2,p)':attack=50:release=100
adrc=transfer='min(p,-60)':attack=2:release=10
Apply dynamic equalization to input audio stream.
A description of the accepted options follows.
Default mode is cut.
Default type is bandpass.
Default type is bell.
Default direction is downward.
Available values are:
Commands
This filter supports the all above options as commands.
Apply dynamic smoothing to input audio stream.
A description of the accepted options follows.
Commands
This filter supports the all above options as commands.
Apply echoing to the input audio.
Echoes are reflected sound and can occur naturally amongst mountains (and sometimes large buildings) when talking or shouting; digital echo effects emulate this behaviour and are often used to help fill out the sound of a single instrument or vocal. The time difference between the original signal and the reflection is the "delay", and the loudness of the reflected signal is the "decay". Multiple echoes can have different delays and decays.
A description of the accepted parameters follows.
Examples
aecho=0.8:0.88:60:0.4
aecho=0.8:0.88:6:0.4
aecho=0.8:0.9:1000:0.3
aecho=0.8:0.9:1000|1800:0.3|0.25
Audio emphasis filter creates or restores material directly taken from LPs or emphased CDs with different filter curves. E.g. to store music on vinyl the signal has to be altered by a filter first to even out the disadvantages of this recording medium. Once the material is played back the inverse filter has to be applied to restore the distortion of the frequency response.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Modify an audio signal according to the specified expressions.
This filter accepts one or more expressions (one for each channel), which are evaluated and used to modify a corresponding audio signal.
It accepts the following parameters:
Each expression in exprs can contain the following constants and functions:
Note: this filter is slow. For faster processing you should use a dedicated filter.
Examples
aeval=val(ch)/2:c=same
aeval=val(0)|-val(1)
An exciter is used to produce high sound that is not present in the original signal. This is done by creating harmonic distortions of the signal which are restricted in range and added to the original signal. An Exciter raises the upper end of an audio signal without simply raising the higher frequencies like an equalizer would do to create a more "crisp" or "brilliant" sound.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Apply fade-in/out effect to input audio.
A description of the accepted parameters follows.
It accepts the following values:
Commands
This filter supports the all above options as commands.
Examples
afade=t=in:ss=0:d=15
afade=t=out:st=875:d=25
Denoise audio samples with FFT.
A description of the accepted parameters follows.
It accepts the following values:
Default value is white noise.
It accepts the following values:
It accepts the following values:
It accepts the following values:
Default value is "none".
Commands
This filter supports the some above mentioned options as commands.
Examples
afftdn=nr=10:nf=-40
afftdn=nr=10:nf=-80:tn=1
asendcmd=0.0 afftdn sn start,asendcmd=0.4 afftdn sn stop,afftdn=nr=20:nf=-40
Apply arbitrary expressions to samples in frequency domain.
Each expression in real and imag can contain the following constants and functions:
It accepts the following values:
Default is "hann".
Examples
afftfilt="'real=re * (1-clip((b/nb)*b,0,1))':imag='im * (1-clip((b/nb)*b,0,1))'"
afftfilt="real='hypot(re,im)*sin(0)':imag='hypot(re,im)*cos(0)':win_size=512:overlap=0.75"
afftfilt="real='hypot(re,im)*cos((random(0)*2-1)*2*3.14)':imag='hypot(re,im)*sin((random(1)*2-1)*2*3.14)':win_size=128:overlap=0.8"
afftfilt="real=re*cos(1)-im*sin(1):imag=re*sin(1)+im*cos(1)"
Apply an arbitrary Finite Impulse Response filter.
This filter is designed for applying long FIR filters, up to 60 seconds long.
It can be used as component for digital crossover filters, room equalization, cross talk cancellation, wavefield synthesis, auralization, ambiophonics, ambisonics and spatialization.
This filter uses the streams higher than first one as FIR coefficients. If the non-first stream holds a single channel, it will be used for all input channels in the first stream, otherwise the number of channels in the non-first stream must be same as the number of channels in the first stream.
It accepts the following parameters:
Set which approach to use for auto gain measurement.
Default value is auto.
Examples
ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
"pan=4C|c0=FL|c1=FL|c2=FR|c3=FR[a];amovie=l_ir.wav[LIR];amovie=r_ir.wav[RIR];[LIR][RIR]amerge[ir];[a][ir]afir=irfmt=input:gtype=gn:irgain=-5dB,pan=stereo|FL<c0+c2|FR<c1+c3"
Set output format constraints for the input audio. The framework will negotiate the most appropriate format to minimize conversions.
It accepts the following parameters:
See the Channel Layout section in the ffmpeg-utils(1) manual for the required syntax.
If a parameter is omitted, all values are allowed.
Force the output to either unsigned 8-bit or signed 16-bit stereo
aformat=sample_fmts=u8|s16:channel_layouts=stereo
Apply frequency shift to input audio samples.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Reduce broadband noise from input samples using Wavelets.
A description of the accepted options follows.
Commands
This filter supports the all above options as commands.
A gate is mainly used to reduce lower parts of a signal. This kind of signal processing reduces disturbing noise between useful signals.
Gating is done by detecting the volume below a chosen level threshold and dividing it by the factor set with ratio. The bottom of the noise floor is set via range. Because an exact manipulation of the signal would cause distortion of the waveform the reduction can be levelled over time. This is done by setting attack and release.
attack determines how long the signal has to fall below the threshold before any reduction will occur and release sets the time the signal has to rise above the threshold to reduce the reduction again. Shorter signals than the chosen attack time will be left untouched.
Commands
This filter supports the all above options as commands.
Apply an arbitrary Infinite Impulse Response filter.
It accepts the following parameters:
Coefficients in "tf" and "sf" format are separated by spaces and are in ascending order.
Coefficients in "zp" format are separated by spaces and order of coefficients doesn't matter. Coefficients in "zp" format are complex numbers with i imaginary unit.
Different coefficients and gains can be provided for every channel, in such case use '|' to separate coefficients or gains. Last provided coefficients will be used for all remaining channels.
Examples
aiir=k=1:z=7.957584807809675810E-1 -2.575128568908332300 3.674839853930788710 -2.57512875289799137 7.957586296317130880E-1:p=1 -2.86950072432325953 3.63022088054647218 -2.28075678147272232 6.361362326477423500E-1:f=tf:r=d
aiir=k=0.79575848078096756:z=0.80918701+0.58773007i 0.80918701-0.58773007i 0.80884700+0.58784055i 0.80884700-0.58784055i:p=0.63892345+0.59951235i 0.63892345-0.59951235i 0.79582691+0.44198673i 0.79582691-0.44198673i:f=zp:r=s
aiir=z=1.3057 0 0 0:p=1.3057 2.3892 2.1860 1:f=sf:r=d
The limiter prevents an input signal from rising over a desired threshold. This limiter uses lookahead technology to prevent your signal from distorting. It means that there is a small delay after the signal is processed. Keep in mind that the delay it produces is the attack time you set.
The filter accepts the following options:
Depending on picked setting it is recommended to upsample input 2x or 4x times with aresample before applying this filter.
Apply a two-pole all-pass filter with central frequency (in Hz) frequency, and filter-width width. An all-pass filter changes the audio's frequency to phase relationship without changing its frequency to amplitude relationship.
The filter accepts the following options:
Commands
This filter supports the following commands:
Loop audio samples.
The filter accepts the following options:
Merge two or more audio streams into a single multi-channel stream.
The filter accepts the following options:
If the channel layouts of the inputs are disjoint, and therefore compatible, the channel layout of the output will be set accordingly and the channels will be reordered as necessary. If the channel layouts of the inputs are not disjoint, the output will have all the channels of the first input then all the channels of the second input, in that order, and the channel layout of the output will be the default value corresponding to the total number of channels.
For example, if the first input is in 2.1 (FL+FR+LF) and the second input is FC+BL+BR, then the output will be in 5.1, with the channels in the following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the first input, b1 is the first channel of the second input).
On the other hand, if both input are in stereo, the output channels will be in the default order: a1, a2, b1, b2, and the channel layout will be arbitrarily set to 4.0, which may or may not be the expected value.
All inputs must have the same sample rate, and format.
If inputs do not have the same duration, the output will stop with the shortest.
Examples
amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
ffmpeg -i input.mkv -filter_complex "[0:1][0:2][0:3][0:4][0:5][0:6] amerge=inputs=6" -c:a pcm_s16le output.mkv
Mixes multiple audio inputs into a single output.
Note that this filter only supports float samples (the amerge and pan audio filters support many formats). If the amix input has integer samples then aresample will be automatically inserted to perform the conversion to float samples.
It accepts the following parameters:
Examples
ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
ffmpeg -i VOCALS -i MUSIC -filter_complex amix=inputs=2:duration=longest:dropout_transition=0:weights="1 0.25":normalize=0 OUTPUT
Commands
This filter supports the following commands:
Multiply first audio stream with second audio stream and store result in output audio stream. Multiplication is done by multiplying each sample from first stream with sample at same position from second stream.
With this element-wise multiplication one can create amplitude fades and amplitude modulations.
High-order parametric multiband equalizer for each channel.
It accepts the following parameters:
Examples
anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
Commands
This filter supports the following commands:
fN is existing filter number, starting from 0, if no such filter is available error is returned. freq set new frequency parameter. width set new width parameter in Hertz. gain set new gain parameter in dB.
Full filter invocation with asendcmd may look like this: asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
Reduce broadband noise in audio samples using Non-Local Means algorithm.
Each sample is adjusted by looking for other samples with similar contexts. This context similarity is defined by comparing their surrounding patches of size p. Patches are searched in an area of r around the sample.
The filter accepts the following options:
It accepts the following values:
Commands
This filter supports the all above options as commands.
Apply Normalized Least-Mean-(Squares|Fourth) algorithm to the first audio stream using the second audio stream.
This adaptive filter is used to mimic a desired filter by finding the filter coefficients that relate to producing the least mean square of the error signal (difference between the desired, 2nd input audio stream and the actual signal, the 1st input audio stream).
A description of the accepted options follows.
Examples
asplit[a][b],[a]adelay=32S|32S[a],[b][a]anlms=order=128:leakage=0.0005:mu=.5:out_mode=o
Commands
This filter supports the same commands as options, excluding option "order".
Pass the audio source unchanged to the output.
Pad the end of an audio stream with silence.
This can be used together with ffmpeg -shortest to extend audio streams to the same length as the video stream.
A description of the accepted options follows.
If neither the pad_len nor the whole_len nor pad_dur nor whole_dur option is set, the filter will add silence to the end of the input stream indefinitely.
Note that for ffmpeg 4.4 and earlier a zero pad_dur or whole_dur also caused the filter to add silence indefinitely.
Examples
apad=pad_len=1024
apad=whole_len=10000
ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
Add a phasing effect to the input audio.
A phaser filter creates series of peaks and troughs in the frequency spectrum. The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
A description of the accepted parameters follows.
Apply phase shift to input audio samples.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Measure Audio Peak Signal-to-Noise Ratio.
This filter takes two audio streams for input, and outputs first audio stream. Results are in dB per channel at end of either input.
Apply Psychoacoustic clipper to input audio stream.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Audio pulsator is something between an autopanner and a tremolo. But it can produce funny stereo effects as well. Pulsator changes the volume of the left and right channel based on a LFO (low frequency oscillator) with different waveforms and shifted phases. This filter have the ability to define an offset between left and right channel. An offset of 0 means that both LFO shapes match each other. The left and right channel are altered equally - a conventional tremolo. An offset of 50% means that the shape of the right channel is exactly shifted in phase (or moved backwards about half of the frequency) - pulsator acts as an autopanner. At 1 both curves match again. Every setting in between moves the phase shift gapless between all stages and produces some "bypassing" sounds with sine and triangle waveforms. The more you set the offset near 1 (starting from the 0.5) the faster the signal passes from the left to the right speaker.
The filter accepts the following options:
Resample the input audio to the specified parameters, using the libswresample library. If none are specified then the filter will automatically convert between its input and output.
This filter is also able to stretch/squeeze the audio data to make it match the timestamps or to inject silence / cut out audio to make it match the timestamps, do a combination of both or do neither.
The filter accepts the syntax [sample_rate:]resampler_options, where sample_rate expresses a sample rate and resampler_options is a list of key=value pairs, separated by ":". See the "Resampler Options" section in the ffmpeg-resampler(1) manual for the complete list of supported options.
Examples
aresample=44100
aresample=async=1000
Reverse an audio clip.
Warning: This filter requires memory to buffer the entire clip, so trimming is suggested.
Examples
atrim=end=5,areverse
Apply Recursive Least Squares algorithm to the first audio stream using the second audio stream.
This adaptive filter is used to mimic a desired filter by recursively finding the filter coefficients that relate to producing the minimal weighted linear least squares cost function of the error signal (difference between the desired, 2nd input audio stream and the actual signal, the 1st input audio stream).
A description of the accepted options follows.
Reduce noise from speech using Recurrent Neural Networks.
This filter accepts the following options:
Commands
This filter supports the all above options as commands.
Measure Audio Signal-to-Distortion Ratio.
This filter takes two audio streams for input, and outputs first audio stream. Results are in dB per channel at end of either input.
Set the number of samples per each output audio frame.
The last output packet may contain a different number of samples, as the filter will flush all the remaining samples when the input audio signals its end.
The filter accepts the following options:
For example, to set the number of per-frame samples to 1234 and disable padding for the last frame, use:
asetnsamples=n=1234:p=0
Set the sample rate without altering the PCM data. This will result in a change of speed and pitch.
The filter accepts the following options:
Show a line containing various information for each input audio frame. The input audio is not modified.
The shown line contains a sequence of key/value pairs of the form key:value.
The following values are shown in the output:
Measure Audio Scaled-Invariant Signal-to-Distortion Ratio.
This filter takes two audio streams for input, and outputs first audio stream. Results are in dB per channel at end of either input.
Apply audio soft clipping.
Soft clipping is a type of distortion effect where the amplitude of a signal is saturated along a smooth curve, rather than the abrupt shape of hard-clipping.
This filter accepts the following options:
It accepts the following values:
Commands
This filter supports the all above options as commands.
Display frequency domain statistical information about the audio channels. Statistics are calculated and stored as metadata for each audio channel and for each audio frame.
It accepts the following option:
It accepts the following values:
Default is "hann".
A list of each metadata key follows:
Automatic Speech Recognition
This filter uses PocketSphinx for speech recognition. To enable compilation of this filter, you need to configure FFmpeg with "--enable-pocketsphinx".
It accepts the following options:
The filter exports recognized speech as the frame metadata "lavfi.asr.text".
Display time domain statistical information about the audio channels. Statistics are calculated and displayed for each audio channel and, where applicable, an overall figure is also given.
It accepts the following option:
Available keys for each channel are: Bit_depth Crest_factor DC_offset Dynamic_range Entropy Flat_factor Max_difference Max_level Mean_difference Min_difference Min_level Noise_floor Noise_floor_count Number_of_Infs Number_of_NaNs Number_of_denormals Peak_count Abs_Peak_count Peak_level RMS_difference RMS_peak RMS_trough Zero_crossings Zero_crossings_rate
and for "Overall": Bit_depth DC_offset Entropy Flat_factor Max_difference Max_level Mean_difference Min_difference Min_level Noise_floor Noise_floor_count Number_of_Infs Number_of_NaNs Number_of_denormals Number_of_samples Peak_count Abs_Peak_count Peak_level RMS_difference RMS_level RMS_peak RMS_trough
For example, a full key looks like "lavfi.astats.1.DC_offset" or "lavfi.astats.Overall.Peak_count".
Read below for the description of the keys.
A description of the measure keys follow:
Boost subwoofer frequencies.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Cut subwoofer frequencies.
This filter allows to set custom, steeper roll off than highpass filter, and thus is able to more attenuate frequency content in stop-band.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Cut super frequencies.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Apply high order Butterworth band-pass filter.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Apply high order Butterworth band-stop filter.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Adjust audio tempo.
The filter accepts exactly one parameter, the audio tempo. If not specified then the filter will assume nominal 1.0 tempo. Tempo must be in the [0.5, 100.0] range.
Note that tempo greater than 2 will skip some samples rather than blend them in. If for any reason this is a concern it is always possible to daisy-chain several instances of atempo to achieve the desired product tempo.
Examples
atempo=0.8
atempo=3
atempo=sqrt(3),atempo=sqrt(3)
Commands
This filter supports the following commands:
Apply spectral tilt filter to audio stream.
This filter apply any spectral roll-off slope over any specified frequency band.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Trim the input so that the output contains one continuous subpart of the input.
It accepts the following parameters:
start, end, and duration are expressed as time duration specifications; see the Time duration section in the ffmpeg-utils(1) manual.
Note that the first two sets of the start/end options and the duration option look at the frame timestamp, while the _sample options simply count the samples that pass through the filter. So start/end_pts and start/end_sample will give different results when the timestamps are wrong, inexact or do not start at zero. Also note that this filter does not modify the timestamps. If you wish to have the output timestamps start at zero, insert the asetpts filter after the atrim filter.
If multiple start or end options are set, this filter tries to be greedy and keep all samples that match at least one of the specified constraints. To keep only the part that matches all the constraints at once, chain multiple atrim filters.
The defaults are such that all the input is kept. So it is possible to set e.g. just the end values to keep everything before the specified time.
Examples:
ffmpeg -i INPUT -af atrim=60:120
ffmpeg -i INPUT -af atrim=end_sample=1000
Calculate normalized windowed cross-correlation between two input audio streams.
Resulted samples are always between -1 and 1 inclusive. If result is 1 it means two input samples are highly correlated in that selected segment. Result 0 means they are not correlated at all. If result is -1 it means two input samples are out of phase, which means they cancel each other.
The filter accepts the following options:
Examples
ffmpeg -i stereo.wav -af channelsplit,axcorrelate=size=1024:algo=fast correlation.wav
Apply a two-pole Butterworth band-pass filter with central frequency frequency, and (3dB-point) band-width width. The csg option selects a constant skirt gain (peak gain = Q) instead of the default: constant 0dB peak gain. The filter roll off at 6dB per octave (20dB per decade).
The filter accepts the following options:
Note that filter delay will be exactly this many samples when set to non-zero value.
Commands
This filter supports the following commands:
Apply a two-pole Butterworth band-reject filter with central frequency frequency, and (3dB-point) band-width width. The filter roll off at 6dB per octave (20dB per decade).
The filter accepts the following options:
Note that filter delay will be exactly this many samples when set to non-zero value.
Commands
This filter supports the following commands:
Boost or cut the bass (lower) frequencies of the audio using a two-pole shelving filter with a response similar to that of a standard hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
The filter accepts the following options:
Note that filter delay will be exactly this many samples when set to non-zero value.
Commands
This filter supports the following commands:
Apply a biquad IIR filter with the given coefficients. Where b0, b1, b2 and a0, a1, a2 are the numerator and denominator coefficients respectively. and channels, c specify which channels to filter, by default all available are filtered.
Commands
This filter supports the following commands:
Note that filter delay will be exactly this many samples when set to non-zero value.
Bauer stereo to binaural transformation, which improves headphone listening of stereo audio records.
To enable compilation of this filter you need to configure FFmpeg with "--enable-libbs2b".
It accepts the following parameters:
Remap input channels to new locations.
It accepts the following parameters:
If no mapping is present, the filter will implicitly map input channels to output channels, preserving indices.
Examples
ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
will create an output WAV file tagged as stereo from the downmix channels of the input.
ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
Split each channel from an input audio stream into a separate output stream.
It accepts the following parameters:
Choosing channels not present in channel layout in the input will result in an error.
Examples
ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
will create an output Matroska file with two audio streams, one containing only the left channel and the other the right channel.
ffmpeg -i in.wav -filter_complex
'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
-map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
side_right.wav
ffmpeg -i in.wav -filter_complex 'channelsplit=channel_layout=5.1:channels=LFE[LFE]'
-map '[LFE]' lfe.wav
Add a chorus effect to the audio.
Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
Chorus resembles an echo effect with a short delay, but whereas with echo the delay is constant, with chorus, it is varied using using sinusoidal or triangular modulation. The modulation depth defines the range the modulated delay is played before or after the delay. Hence the delayed sound will sound slower or faster, that is the delayed sound tuned around the original one, like in a chorus where some vocals are slightly off key.
It accepts the following parameters:
Examples
chorus=0.7:0.9:55:0.4:0.25:2
chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
chorus=0.5:0.9:50|60|40:0.4|0.32|0.3:0.25|0.4|0.3:2|2.3|1.3
Compress or expand the audio's dynamic range.
It accepts the following parameters:
The input values must be in strictly increasing order but the transfer function does not have to be monotonically rising. The point "0/0" is assumed but may be overridden (by "0/out-dBn"). Typical values for the transfer function are "-70/-70|-60/-20|1/0".
Examples
compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
Another example for audio with whisper and explosion parts:
compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
compand=attacks=0:points=-80/-169|-54/-80|-49.5/-64.6|-41.1/-41.1|-25.8/-15|-10.8/-4.5|0/0|20/8.3
compand=attacks=0:points=-80/-80|-6/-6|20/-6
compand=attacks=0:points=-80/-80|-12/-12|20/-12
compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
Compensation Delay Line is a metric based delay to compensate differing positions of microphones or speakers.
For example, you have recorded guitar with two microphones placed in different locations. Because the front of sound wave has fixed speed in normal conditions, the phasing of microphones can vary and depends on their location and interposition. The best sound mix can be achieved when these microphones are in phase (synchronized). Note that a distance of ~30 cm between microphones makes one microphone capture the signal in antiphase to the other microphone. That makes the final mix sound moody. This filter helps to solve phasing problems by adding different delays to each microphone track and make them synchronized.
The best result can be reached when you take one track as base and synchronize other tracks one by one with it. Remember that synchronization/delay tolerance depends on sample rate, too. Higher sample rates will give more tolerance.
The filter accepts the following parameters:
Commands
This filter supports the all above options as commands.
Apply headphone crossfeed filter.
Crossfeed is the process of blending the left and right channels of stereo audio recording. It is mainly used to reduce extreme stereo separation of low frequencies.
The intent is to produce more speaker like sound to the listener.
The filter accepts the following options:
Note that filter delay will be exactly this many samples when set to non-zero value.
Commands
This filter supports the all above options as commands.
Simple algorithm for audio noise sharpening.
This filter linearly increases differences betweeen each audio sample.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Apply a DC shift to the audio.
This can be useful to remove a DC offset (caused perhaps by a hardware problem in the recording chain) from the audio. The effect of a DC offset is reduced headroom and hence volume. The astats filter can be used to determine if a signal has a DC offset.
Apply de-essing to the audio samples.
It accepts the following values:
Enhance dialogue in stereo audio.
This filter accepts stereo input and produce surround (3.0) channels output. The newly produced front center channel have enhanced speech dialogue originally available in both stereo channels. This filter outputs front left and front right channels same as available in stereo input.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Measure audio dynamic range.
DR values of 14 and higher is found in very dynamic material. DR of 8 to 13 is found in transition material. And anything less that 8 have very poor dynamics and is very compressed.
The filter accepts the following options:
Dynamic Audio Normalizer.
This filter applies a certain amount of gain to the input audio in order to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in contrast to more "simple" normalization algorithms, the Dynamic Audio Normalizer *dynamically* re-adjusts the gain factor to the input audio. This allows for applying extra gain to the "quiet" sections of the audio while avoiding distortions or clipping the "loud" sections. In other words: The Dynamic Audio Normalizer will "even out" the volume of quiet and loud sections, in the sense that the volume of each section is brought to the same target level. Note, however, that the Dynamic Audio Normalizer achieves this goal *without* applying "dynamic range compressing". It will retain 100% of the dynamic range *within* each section of the audio file.
The expression can contain the following constants:
Commands
This filter supports the all above options as commands.
Make audio easier to listen to on headphones.
This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio so that when listened to on headphones the stereo image is moved from inside your head (standard for headphones) to outside and in front of the listener (standard for speakers).
Ported from SoX.
Apply a two-pole peaking equalisation (EQ) filter. With this filter, the signal-level at and around a selected frequency can be increased or decreased, whilst (unlike bandpass and bandreject filters) that at all other frequencies is unchanged.
In order to produce complex equalisation curves, this filter can be given several times, each with a different central frequency.
The filter accepts the following options:
Note that filter delay will be exactly this many samples when set to non-zero value.
Examples
equalizer=f=1000:t=h:width=200:g=-10
equalizer=f=1000:t=q:w=1:g=2,equalizer=f=100:t=q:w=2:g=-5
Commands
This filter supports the following commands:
Linearly increases the difference between left and right channels which adds some sort of "live" effect to playback.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Apply FIR Equalization using arbitrary frequency response.
The filter accepts the following option:
and functions:
This option is also available as command. Default is gain_interpolate(f).
This option is also available as command.
Examples
firequalizer=gain='if(lt(f,1000), 0, -INF)'
firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
firequalizer=delay=0.1:fixed=on:zero_phase=on
firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
:gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
Apply a flanging effect to the audio.
The filter accepts the following options:
Apply Haas effect to audio.
Note that this makes most sense to apply on mono signals. With this filter applied to mono signals it give some directionality and stretches its stereo image.
The filter accepts the following options:
Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with embedded HDCD codes is expanded into a 20-bit PCM stream.
The filter supports the Peak Extend and Low-level Gain Adjustment features of HDCD, and detects the Transient Filter flag.
ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
When using the filter with wav, note the default encoding for wav is 16-bit, so the resulting 20-bit stream will be truncated back to 16-bit. Use something like -acodec pcm_s24le after the filter to get 24-bit PCM output.
ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
ffmpeg -i HDCD16.wav -af hdcd -c:a pcm_s24le OUT24.wav
The filter accepts the following options:
"analyze_mode=pe:force_pe=true" can be used to see all samples above the PE level.
Modes are:
Apply head-related transfer functions (HRTFs) to create virtual loudspeakers around the user for binaural listening via headphones. The HRIRs are provided via additional streams, for each channel one stereo input stream is needed.
The filter accepts the following options:
Examples
ffmpeg -i input.wav
-filter_complex "amovie=azi_270_ele_0_DFC.wav[sr];amovie=azi_90_ele_0_DFC.wav[sl];amovie=azi_225_ele_0_DFC.wav[br];amovie=azi_135_ele_0_DFC.wav[bl];amovie=azi_0_ele_0_DFC.wav,asplit[fc][lfe];amovie=azi_35_ele_0_DFC.wav[fl];amovie=azi_325_ele_0_DFC.wav[fr];[0:a][fl][fr][fc][lfe][bl][br][sl][sr]headphone=FL|FR|FC|LFE|BL|BR|SL|SR"
output.wav
ffmpeg -i input.wav -filter_complex "amovie=minp.wav[hrirs];[0:a][hrirs]headphone=map=FL|FR|FC|LFE|BL|BR|SL|SR:hrir=multich"
output.wav
Apply a high-pass filter with 3dB point frequency. The filter can be either single-pole, or double-pole (the default). The filter roll off at 6dB per pole per octave (20dB per pole per decade).
The filter accepts the following options:
Note that filter delay will be exactly this many samples when set to non-zero value.
Commands
This filter supports the following commands:
Join multiple input streams into one multi-channel stream.
It accepts the following parameters:
The filter will attempt to guess the mappings when they are not specified explicitly. It does so by first trying to find an unused matching input channel and if that fails it picks the first unused input channel.
Join 3 inputs (with properly set channel layouts):
ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
Build a 5.1 output from 6 single-channel streams:
ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
'join=inputs=6:channel_layout=5.1:map=0.0-FL|1.0-FR|2.0-FC|3.0-SL|4.0-SR|5.0-LFE'
out
Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
To enable compilation of this filter you need to configure FFmpeg with "--enable-ladspa".
Examples
ladspa=file=amp
ladspa=f=vcf:p=vcf_notch:c=help
ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
ladspa=file=tap_reverb:tap_reverb
ladspa=file=cmt:noise_source_white:c=c0=.2
ladspa=file=caps:Click:c=c1=20'
ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
ladspa=caps:Narrower
ladspa=caps:White:.2
ladspa=caps:Fractal:c=c1=1
ladspa=vlevel-ladspa:vlevel_mono
Commands
This filter supports the following commands:
If the specified value is not valid, it is ignored and prior one is kept.
EBU R128 loudness normalization. Includes both dynamic and linear normalization modes. Support for both single pass (livestreams, files) and double pass (files) modes. This algorithm can target IL, LRA, and maximum true peak. In dynamic mode, to accurately detect true peaks, the audio stream will be upsampled to 192 kHz. Use the "-ar" option or "aresample" filter to explicitly set an output sample rate.
The filter accepts the following options:
Apply a low-pass filter with 3dB point frequency. The filter can be either single-pole or double-pole (the default). The filter roll off at 6dB per pole per octave (20dB per pole per decade).
The filter accepts the following options:
Note that filter delay will be exactly this many samples when set to non-zero value.
Examples
lowpass=c=LFE
Commands
This filter supports the following commands:
Load a LV2 (LADSPA Version 2) plugin.
To enable compilation of this filter you need to configure FFmpeg with "--enable-lv2".
Examples
lv2=p=http\\\\://calf.sourceforge.net/plugins/BassEnhancer:c=amount=2
lv2=p=http\\\\://calf.sourceforge.net/plugins/Vinyl:c=drone=0.2|aging=0.5
lv2=p=http\\\\://www.openavproductions.com/artyfx#bitta:c=crush=0.3
Commands
This filter supports all options that are exported by plugin as commands.
Multiband Compress or expand the audio's dynamic range.
The input audio is divided into bands using 4th order Linkwitz-Riley IIRs. This is akin to the crossover of a loudspeaker, and results in flat frequency response when absent compander action.
It accepts the following parameters:
Mix channels with specific gain levels. The filter accepts the output channel layout followed by a set of channels definitions.
This filter is also designed to efficiently remap the channels of an audio stream.
The filter accepts parameters of the form: "l|outdef|outdef|..."
If the `=' in a channel specification is replaced by `<', then the gains for that specification will be renormalized so that the total is 1, thus avoiding clipping noise.
Mixing examples
For example, if you want to down-mix from stereo to mono, but with a bigger factor for the left channel:
pan=1c|c0=0.9*c0+0.1*c1
A customized down-mix to stereo that works automatically for 3-, 4-, 5- and 7-channels surround:
pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
Note that ffmpeg integrates a default down-mix (and up-mix) system that should be preferred (see "-ac" option) unless you have very specific needs.
Remapping examples
The channel remapping will be effective if, and only if:
If all these conditions are satisfied, the filter will notify the user ("Pure channel mapping detected"), and use an optimized and lossless method to do the remapping.
For example, if you have a 5.1 source and want a stereo audio stream by dropping the extra channels:
pan="stereo| c0=FL | c1=FR"
Given the same source, you can also switch front left and front right channels and keep the input channel layout:
pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
If the input is a stereo audio stream, you can mute the front left channel (and still keep the stereo channel layout) with:
pan="stereo|c1=c1"
Still with a stereo audio stream input, you can copy the right channel in both front left and right:
pan="stereo| c0=FR | c1=FR"
ReplayGain scanner filter. This filter takes an audio stream as an input and outputs it unchanged. At end of filtering it displays "track_gain" and "track_peak".
The filter accepts the following exported read-only options:
Convert the audio sample format, sample rate and channel layout. It is not meant to be used directly.
Apply time-stretching and pitch-shifting with librubberband.
To enable compilation of this filter, you need to configure FFmpeg with "--enable-librubberband".
The filter accepts the following options:
Commands
This filter supports the following commands:
This filter acts like normal compressor but has the ability to compress detected signal using second input signal. It needs two input streams and returns one output stream. First input stream will be processed depending on second stream signal. The filtered signal then can be filtered with other filters in later stages of processing. See pan and amerge filter.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Examples
ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
A sidechain gate acts like a normal (wideband) gate but has the ability to filter the detected signal before sending it to the gain reduction stage. Normally a gate uses the full range signal to detect a level above the threshold. For example: If you cut all lower frequencies from your sidechain signal the gate will decrease the volume of your track only if not enough highs appear. With this technique you are able to reduce the resonation of a natural drum or remove "rumbling" of muted strokes from a heavily distorted guitar. It needs two input streams and returns one output stream. First input stream will be processed depending on second stream signal.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Detect silence in an audio stream.
This filter logs a message when it detects that the input audio volume is less or equal to a noise tolerance value for a duration greater or equal to the minimum detected noise duration.
The printed times and duration are expressed in seconds. The "lavfi.silence_start" or "lavfi.silence_start.X" metadata key is set on the first frame whose timestamp equals or exceeds the detection duration and it contains the timestamp of the first frame of the silence.
The "lavfi.silence_duration" or "lavfi.silence_duration.X" and "lavfi.silence_end" or "lavfi.silence_end.X" metadata keys are set on the first frame after the silence. If mono is enabled, and each channel is evaluated separately, the ".X" suffixed keys are used, and "X" corresponds to the channel number.
The filter accepts the following options:
Examples
silencedetect=n=-50dB:d=5
ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
Remove silence from the beginning, middle or end of the audio.
The filter accepts the following options:
Default value is "rms".
Defaults value is "write".
Examples
silenceremove=start_periods=1:start_duration=5:start_threshold=0.02
silenceremove=stop_periods=-1:stop_duration=1:stop_threshold=-90dB
silenceremove=window=0:detection=peak:stop_mode=all:start_mode=all:stop_periods=-1:stop_threshold=0
silenceremove=stop_periods=-2:stop_duration=1:stop_threshold=-90dB
silenceremove=stop_periods=-2:stop_duration=1:stop_threshold=-90dB:stop_silence=0.5
silenceremove=stop_periods=-2:stop_duration=1:stop_threshold=-90dB:stop_silence=0.5:start_periods=1:start_duration=1:start_silence=1.5:stop_threshold=-90dB
Commands
This filter supports some above options as commands.
SOFAlizer uses head-related transfer functions (HRTFs) to create virtual loudspeakers around the user for binaural listening via headphones (audio formats up to 9 channels supported). The HRTFs are stored in SOFA files (see <http://www.sofacoustics.org/> for a database). SOFAlizer is developed at the Acoustics Research Institute (ARI) of the Austrian Academy of Sciences.
To enable compilation of this filter you need to configure FFmpeg with "--enable-libmysofa".
The filter accepts the following options:
Examples
sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
"sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
Speech Normalizer.
This filter expands or compresses each half-cycle of audio samples (local set of samples all above or all below zero and between two nearest zero crossings) depending on threshold value, so audio reaches target peak value under conditions controlled by below options.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Examples
speechnorm=e=3:r=0.00001:l=1
speechnorm=e=6.25:r=0.00001:l=1
speechnorm=e=12.5:r=0.0001:l=1
speechnorm=e=25:r=0.0001:l=1
speechnorm=e=50:r=0.0001:l=1
This filter has some handy utilities to manage stereo signals, for converting M/S stereo recordings to L/R signal while having control over the parameters or spreading the stereo image of master track.
The filter accepts the following options:
Can be one of the following:
Commands
This filter supports the all above options as commands.
Examples
stereotools=mlev=0.015625
"stereotools=mode=ms>lr"
This filter enhance the stereo effect by suppressing signal common to both channels and by delaying the signal of left into right and vice versa, thereby widening the stereo effect.
The filter accepts the following options:
Commands
This filter supports the all above options except "delay" as commands.
Apply 18 band equalizer.
The filter accepts the following options:
Apply audio surround upmix filter.
This filter allows to produce multichannel output from audio stream.
The filter accepts the following options:
See the Channel Layout section in the ffmpeg-utils(1) manual for the required syntax.
See the Channel Layout section in the ffmpeg-utils(1) manual for the required syntax.
It accepts the following values:
Default is "hann".
Boost or cut the lower frequencies and cut or boost higher frequencies of the audio using a two-pole shelving filter with a response similar to that of a standard hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
The filter accepts the following options:
Note that filter delay will be exactly this many samples when set to non-zero value.
Commands
This filter supports some options as commands.
Boost or cut treble (upper) frequencies of the audio using a two-pole shelving filter with a response similar to that of a standard hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
The filter accepts the following options:
Note that filter delay will be exactly this many samples when set to non-zero value.
Commands
This filter supports the following commands:
Sinusoidal amplitude modulation.
The filter accepts the following options:
Sinusoidal phase modulation.
The filter accepts the following options:
Apply audio Virtual Bass filter.
This filter accepts stereo input and produce stereo with LFE (2.1) channels output. The newly produced LFE channel have enhanced virtual bass originally obtained from both stereo channels. This filter outputs front left and front right channels unchanged as available in stereo input.
The filter accepts the following options:
Adjust the input audio volume.
It accepts the following parameters:
Output values are clipped to the maximum value.
The output audio volume is given by the relation:
<output_volume> = <volume> * <input_volume>
The default value for volume is "1.0".
It determines which input sample formats will be allowed, which affects the precision of the volume scaling.
Default value for replaygain_preamp is 0.0.
Default value for replaygain_noclip is 1.
It accepts the following values:
Default value is once.
The volume expression can contain the following parameters.
Note that when eval is set to once only the sample_rate and tb variables are available, all other variables will evaluate to NAN.
Commands
This filter supports the following commands:
If the specified expression is not valid, it is kept at its current value.
Examples
volume=volume=0.5
volume=volume=1/2
volume=volume=-6.0206dB
In all the above example the named key for volume can be omitted, for example like in:
volume=0.5
volume=volume=6dB:precision=fixed
volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
Detect the volume of the input video.
The filter has no parameters. It supports only 16-bit signed integer samples, so the input will be converted when needed. Statistics about the volume will be printed in the log when the input stream end is reached.
In particular it will show the mean volume (root mean square), maximum volume (on a per-sample basis), and the beginning of a histogram of the registered volume values (from the maximum value to a cumulated 1/1000 of the samples).
All volumes are in decibels relative to the maximum PCM value.
Examples
Here is an excerpt of the output:
[Parsed_volumedetect_0 0xa23120] mean_volume: -27 dB
[Parsed_volumedetect_0 0xa23120] max_volume: -4 dB
[Parsed_volumedetect_0 0xa23120] histogram_4db: 6
[Parsed_volumedetect_0 0xa23120] histogram_5db: 62
[Parsed_volumedetect_0 0xa23120] histogram_6db: 286
[Parsed_volumedetect_0 0xa23120] histogram_7db: 1042
[Parsed_volumedetect_0 0xa23120] histogram_8db: 2551
[Parsed_volumedetect_0 0xa23120] histogram_9db: 4609
[Parsed_volumedetect_0 0xa23120] histogram_10db: 8409
It means that:
In other words, raising the volume by +4 dB does not cause any clipping, raising it by +5 dB causes clipping for 6 samples, etc.
Below is a description of the currently available audio sources.
Buffer audio frames, and make them available to the filter chain.
This source is mainly intended for a programmatic use, in particular through the interface defined in libavfilter/buffersrc.h.
It accepts the following parameters:
Examples
abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
will instruct the source to accept planar 16bit signed stereo at 44100Hz. Since the sample format with name "s16p" corresponds to the number 6 and the "stereo" channel layout corresponds to the value 0x3, this is equivalent to:
abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
Generate an audio signal specified by an expression.
This source accepts in input one or more expressions (one for each channel), which are evaluated and used to generate a corresponding audio signal.
This source accepts the following options:
If not specified, or the expressed duration is negative, the audio is supposed to be generated forever.
Each expression in exprs can contain the following constants:
Examples
aevalsrc=0
aevalsrc="sin(440*2*PI*t):s=8000"
aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
aevalsrc="-2+random(0)"
aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
Generate a fractional delay FIR coefficients.
The resulting stream can be used with afir filter for filtering the audio signal.
The filter accepts the following options:
Generate a FIR equalizer coefficients.
The resulting stream can be used with afir filter for filtering the audio signal.
The filter accepts the following options:
Available presets are:
Generate a FIR coefficients using frequency sampling method.
The resulting stream can be used with afir filter for filtering the audio signal.
The filter accepts the following options:
The null audio source, return unprocessed audio frames. It is mainly useful as a template and to be employed in analysis / debugging tools, or as the source for filters which ignore the input data (for example the sox synth filter).
This source accepts the following options:
Check the channel_layout_map definition in libavutil/channel_layout.c for the mapping between strings and channel layout values.
If not specified, or the expressed duration is negative, the audio is supposed to be generated forever.
Examples
anullsrc=r=48000:cl=4
anullsrc=r=48000:cl=mono
All the parameters need to be explicitly defined.
Synthesize a voice utterance using the libflite library.
To enable compilation of this filter you need to configure FFmpeg with "--enable-libflite".
Note that versions of the flite library prior to 2.0 are not thread-safe.
The filter accepts the following options:
Examples
flite=textfile=speech.txt
flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
For more information about libflite, check: <http://www.festvox.org/flite/>
Generate a noise audio signal.
The filter accepts the following options:
Examples
anoisesrc=d=60:c=pink:r=44100:a=0.5
Generate odd-tap Hilbert transform FIR coefficients.
The resulting stream can be used with afir filter for phase-shifting the signal by 90 degrees.
This is used in many matrix coding schemes and for analytic signal generation. The process is often written as a multiplication by i (or j), the imaginary unit.
The filter accepts the following options:
Generate a sinc kaiser-windowed low-pass, high-pass, band-pass, or band-reject FIR coefficients.
The resulting stream can be used with afir filter for filtering the audio signal.
The filter accepts the following options:
Generate an audio signal made of a sine wave with amplitude 1/8.
The audio signal is bit-exact.
The filter accepts the following options:
The expression can contain the following constants:
Default is 1024.
Examples
sine
sine=220:4:d=5
sine=f=220:b=4:d=5
sine=frequency=220:beep_factor=4:duration=5
sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
Below is a description of the currently available audio sinks.
Buffer audio frames, and make them available to the end of filter chain.
This sink is mainly intended for programmatic use, in particular through the interface defined in libavfilter/buffersink.h or the options system.
It accepts a pointer to an AVABufferSinkContext structure, which defines the incoming buffers' formats, to be passed as the opaque parameter to "avfilter_init_filter" for initialization.
Null audio sink; do absolutely nothing with the input audio. It is mainly useful as a template and for use in analysis / debugging tools.
When you configure your FFmpeg build, you can disable any of the existing filters using "--disable-filters". The configure output will show the video filters included in your build.
Below is a description of the currently available video filters.
Mark a region of interest in a video frame.
The frame data is passed through unchanged, but metadata is attached to the frame indicating regions of interest which can affect the behaviour of later encoding. Multiple regions can be marked by applying the filter multiple times.
The parameters x, y, w and h are expressions, and may contain the following variables:
This must be a real value in the range -1 to +1. A value of zero indicates no quality change. A negative value asks for better quality (less quantisation), while a positive value asks for worse quality (greater quantisation).
The range is calibrated so that the extreme values indicate the largest possible offset - if the rest of the frame is encoded with the worst possible quality, an offset of -1 indicates that this region should be encoded with the best possible quality anyway. Intermediate values are then interpolated in some codec-dependent way.
For example, in 10-bit H.264 the quantisation parameter varies between -12 and 51. A typical qoffset value of -1/10 therefore indicates that this region should be encoded with a QP around one-tenth of the full range better than the rest of the frame. So, if most of the frame were to be encoded with a QP of around 30, this region would get a QP of around 24 (an offset of approximately -1/10 * (51 - -12) = -6.3). An extreme value of -1 would indicate that this region should be encoded with the best possible quality regardless of the treatment of the rest of the frame - that is, should be encoded at a QP of -12.
Examples
addroi=iw/4:ih/4:iw/2:ih/2:-1/10
addroi=0:0:100:ih:+1/5
Extract the alpha component from the input as a grayscale video. This is especially useful with the alphamerge filter.
Add or replace the alpha component of the primary input with the grayscale value of a second input. This is intended for use with alphaextract to allow the transmission or storage of frame sequences that have alpha in a format that doesn't support an alpha channel.
For example, to reconstruct full frames from a normal YUV-encoded video and a separate video created with alphaextract, you might use:
movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
Amplify differences between current pixel and pixels of adjacent frames in same pixel location.
This filter accepts the following options:
Commands
This filter supports the following commands that corresponds to option of same name:
Same as the subtitles filter, except that it doesn't require libavcodec and libavformat to work. On the other hand, it is limited to ASS (Advanced Substation Alpha) subtitles files.
This filter accepts the following option in addition to the common options from the subtitles filter:
Available values are:
The default is "auto".
Apply an Adaptive Temporal Averaging Denoiser to the video input.
The filter accepts the following options:
Threshold A is designed to react on abrupt changes in the input signal and threshold B is designed to react on continuous changes in the input signal.
Parallel can be faster then serial, while other way around is never true. Parallel will abort early on first change being greater then thresholds, while serial will continue processing other side of frames if they are equal or below thresholds.
Commands
This filter supports same commands as options except option "s". The command accepts the same syntax of the corresponding option.
Apply average blur filter.
The filter accepts the following options:
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Turns a static background into transparency.
The filter accepts the following option:
Commands
This filter supports the all above options as commands.
Compute the bounding box for the non-black pixels in the input frame luma plane.
This filter computes the bounding box containing all the pixels with a luma value greater than the minimum allowed value. The parameters describing the bounding box are printed on the filter log.
The filter accepts the following option:
Commands
This filter supports the all above options as commands.
Apply bilateral filter, spatial smoothing while preserving edges.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
CUDA accelerated bilateral filter, an edge preserving filter. This filter is mathematically accurate thanks to the use of GPU acceleration. For best output quality, use one to one chroma subsampling, i.e. yuv444p format.
The filter accepts the following options:
Examples
./ffmpeg -v verbose \
-hwaccel cuda -hwaccel_output_format cuda -i input.mp4 \
-init_hw_device cuda \
-filter_complex \
" \
[0:v]scale_cuda=format=yuv444p[scaled_video];
[scaled_video]bilateral_cuda=window_size=9:sigmaS=3.0:sigmaR=50.0" \
-an -sn -c:v h264_nvenc -cq 20 out.mp4
Show and measure bit plane noise.
The filter accepts the following options:
Detect video intervals that are (almost) completely black. Can be useful to detect chapter transitions, commercials, or invalid recordings.
The filter outputs its detection analysis to both the log as well as frame metadata. If a black segment of at least the specified minimum duration is found, a line with the start and end timestamps as well as duration is printed to the log with level "info". In addition, a log line with level "debug" is printed per frame showing the black amount detected for that frame.
The filter also attaches metadata to the first frame of a black segment with key "lavfi.black_start" and to the first frame after the black segment ends with key "lavfi.black_end". The value is the frame's timestamp. This metadata is added regardless of the minimum duration specified.
The filter accepts the following options:
Default value is 2.0.
<nb_black_pixels> / <nb_pixels>
for which a picture is considered black. Default value is 0.98.
The threshold expresses the maximum pixel luma value for which a pixel is considered "black". The provided value is scaled according to the following equation:
<absolute_threshold> = <luma_minimum_value> + <pixel_black_th> * <luma_range_size>
luma_range_size and luma_minimum_value depend on the input video format, the range is [0-255] for YUV full-range formats and [16-235] for YUV non full-range formats.
Default value is 0.10.
The following example sets the maximum pixel threshold to the minimum value, and detects only black intervals of 2 or more seconds:
blackdetect=d=2:pix_th=0.00
Detect frames that are (almost) completely black. Can be useful to detect chapter transitions or commercials. Output lines consist of the frame number of the detected frame, the percentage of blackness, the position in the file if known or -1 and the timestamp in seconds.
In order to display the output lines, you need to set the loglevel at least to the AV_LOG_INFO value.
This filter exports frame metadata "lavfi.blackframe.pblack". The value represents the percentage of pixels in the picture that are below the threshold value.
It accepts the following parameters:
Blend two video frames into each other.
The "blend" filter takes two input streams and outputs one stream, the first input is the "top" layer and second input is "bottom" layer. By default, the output terminates when the longest input terminates.
The "tblend" (time blend) filter takes two consecutive frames from one single stream, and outputs the result obtained by blending the new frame on top of the old frame.
A description of the accepted options follows.
Available values for component modes are:
The expressions can use the following variables:
The "blend" filter also supports the framesync options.
Examples
blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
blend=all_expr='A*(X/W)+B*(1-X/W)'
blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
blend=all_expr='if(gte(N*SW+X,W),A,B)'
blend=all_expr='if(gte(Y-N*SH,0),A,B)'
blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
blend=all_expr='if(gt(X,Y*(W/H)),A,B)'
tblend=all_mode=grainextract
Commands
This filter supports same commands as options.
Determines blockiness of frames without altering the input frames.
Based on Remco Muijs and Ihor Kirenko: "A no-reference blocking artifact measure for adaptive video processing." 2005 13th European signal processing conference.
The filter accepts the following options:
Examples
blockdetect=period_min=8:period_max=32:planes=1
Determines blurriness of frames without altering the input frames.
Based on Marziliano, Pina, et al. "A no-reference perceptual blur metric." Allows for a block-based abbreviation.
The filter accepts the following options:
The high threshold selects the "strong" edge pixels, which are then connected through 8-connectivity with the "weak" edge pixels selected by the low threshold.
low and high threshold values must be chosen in the range [0,1], and low should be lesser or equal to high.
Default value for low is "20/255", and default value for high is "50/255".
Examples
blurdetect=block_width=32:block_height=32:block_pct=80
Denoise frames using Block-Matching 3D algorithm.
The filter accepts the following options.
Examples
bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic
bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic:planes=1
split[a][b],[a]bm3d=sigma=3:block=4:bstep=2:group=1:estim=basic[a],[b][a]bm3d=sigma=3:block=4:bstep=2:group=16:estim=final:ref=1
split[a][b],[a]nlmeans=s=3:r=7:p=3[a],[b][a]bm3d=sigma=3:block=4:bstep=2:group=16:estim=final:ref=1
Apply a boxblur algorithm to the input video.
It accepts the following parameters:
A description of the accepted options follows.
The radius value must be a non-negative number, and must not be greater than the value of the expression "min(w,h)/2" for the luma and alpha planes, and of "min(cw,ch)/2" for the chroma planes.
Default value for luma_radius is "2". If not specified, chroma_radius and alpha_radius default to the corresponding value set for luma_radius.
The expressions can contain the following constants:
Default value for luma_power is 2. If not specified, chroma_power and alpha_power default to the corresponding value set for luma_power.
A value of 0 will disable the effect.
Examples
boxblur=luma_radius=2:luma_power=1
boxblur=2:1
boxblur=2:1:cr=0:ar=0
boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
Deinterlace the input video ("bwdif" stands for "Bob Weaver Deinterlacing Filter").
Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic interpolation algorithms. It accepts the following parameters:
The default value is "send_field".
The default value is "auto". If the interlacing is unknown or the decoder does not export this information, top field first will be assumed.
The default value is "all".
Deinterlace the input video using the bwdif algorithm, but implemented in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec and/or nvenc.
It accepts the following parameters:
The default value is "send_field".
The default value is "auto". If the interlacing is unknown or the decoder does not export this information, top field first will be assumed.
The default value is "all".
Repack CEA-708 closed captioning side data
This filter fixes various issues seen with commerical encoders related to upstream malformed CEA-708 payloads, specifically incorrect number of tuples (wrong cc_count for the target FPS), and incorrect ordering of tuples (i.e. the CEA-608 tuples are not at the first entries in the payload).
Apply Contrast Adaptive Sharpen filter to video stream.
The filter accepts the following options:
Commands
This filter supports same commands as options.
Remove all color information for all colors except for certain one.
The filter accepts the following options:
Literal colors like "green" or "red" don't make sense with this enabled anymore. This can be used to pass exact YUV values as hexadecimal numbers.
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
YUV colorspace color/chroma keying.
The filter accepts the following options:
0.01 matches only the exact key color, while 1.0 matches everything.
0.0 makes pixels either fully transparent, or not transparent at all.
Higher values result in semi-transparent pixels, with a higher transparency the more similar the pixels color is to the key color.
Literal colors like "green" or "red" don't make sense with this enabled anymore. This can be used to pass exact YUV values as hexadecimal numbers.
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Examples
ffmpeg -i input.png -vf chromakey=green out.png
ffmpeg -f lavfi -i color=c=black:s=1280x720 -i video.mp4 -shortest -filter_complex "[1:v]chromakey=0x70de77:0.1:0.2[ckout];[0:v][ckout]overlay[out]" -map "[out]" output.mkv
CUDA accelerated YUV colorspace color/chroma keying.
This filter works like normal chromakey filter but operates on CUDA frames. for more details and parameters see chromakey.
Examples
./ffmpeg \
-hwaccel cuda -hwaccel_output_format cuda -i input_green.mp4 \
-hwaccel cuda -hwaccel_output_format cuda -i base_video.mp4 \
-init_hw_device cuda \
-filter_complex \
" \
[0:v]chromakey_cuda=0x25302D:0.1:0.12:1[overlay_video]; \
[1:v]scale_cuda=format=yuv420p[base]; \
[base][overlay_video]overlay_cuda" \
-an -sn -c:v h264_nvenc -cq 20 output.mp4
./ffmpeg -init_hw_device cuda=cuda -filter_hw_device cuda \
-f lavfi -i color=size=800x600:color=white,format=yuv420p \
-f lavfi -i yuvtestsrc=size=200x200,format=yuv420p \
-filter_complex \
" \
[0]hwupload[under]; \
[1]hwupload,chromakey_cuda=green:0.1:0.12[over]; \
[under][over]overlay_cuda" \
-c:v hevc_nvenc -cq 18 -preset slow output.mp4
Reduce chrominance noise.
The filter accepts the following options:
Default distance type is manhattan.
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
Shift chroma pixels horizontally and/or vertically.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Display CIE color diagram with pixels overlaid onto it.
The filter accepts the following options:
See "system" option for available values.
Visualize information exported by some codecs.
Some codecs can export information through frames using side-data or other means. For example, some MPEG based codecs export motion vectors through the export_mvs flag in the codec flags2 option.
The filter accepts the following option:
Available flags for mv are:
Available flags for mv_type are:
Available flags for frame_type are:
Examples
ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
Modify intensity of primary colors (red, green and blue) of input frames.
The filter allows an input frame to be adjusted in the shadows, midtones or highlights regions for the red-cyan, green-magenta or blue-yellow balance.
A positive adjustment value shifts the balance towards the primary color, a negative value towards the complementary color.
The filter accepts the following options:
Allowed ranges for options are "[-1.0, 1.0]". Defaults are 0.
Examples
colorbalance=rs=.3
Commands
This filter supports the all above options as commands.
Adjust color contrast between RGB components.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Adjust color white balance selectively for blacks and whites. This filter operates in YUV colorspace.
The filter accepts the following options:
Possible values are:
Default value is "manual".
Commands
This filter supports the all above options as commands.
Adjust video input frames by re-mixing color channels.
This filter modifies a color channel by adding the values associated to the other channels of the same pixels. For example if the value to modify is red, the output value will be:
<red>=<red>*<rr> + <blue>*<rb> + <green>*<rg> + <alpha>*<ra>
The filter accepts the following options:
Allowed ranges for options are "[-2.0, 2.0]".
Examples
colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
Commands
This filter supports the all above options as commands.
Overlay a solid color on the video stream.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
RGB colorspace color keying. This filter operates on 8-bit RGB format frames by setting the alpha component of each pixel which falls within the similarity radius of the key color to 0. The alpha value for pixels outside the similarity radius depends on the value of the blend option.
The filter accepts the following options:
Examples
ffmpeg -i input.png -vf colorkey=green out.png
ffmpeg -i background.png -i video.mp4 -filter_complex "[1:v]colorkey=0x3BBD1E:0.3:0.2[ckout];[0:v][ckout]overlay[out]" -map "[out]" output.flv
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Remove all color information for all RGB colors except for certain one.
The filter accepts the following options:
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Adjust video input frames using levels.
The filter accepts the following options:
Input levels are used to lighten highlights (bright tones), darken shadows (dark tones), change the balance of bright and dark tones.
Output levels allows manual selection of a constrained output level range.
Examples
colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
colorlevels=romin=0.5:gomin=0.5:bomin=0.5
Commands
This filter supports the all above options as commands.
Apply custom color maps to video stream.
This filter needs three input video streams. First stream is video stream that is going to be filtered out. Second and third video stream specify color patches for source color to target color mapping.
The filter accepts the following options:
The accepted values are:
Default is "euclidean".
Convert color matrix.
The filter accepts the following options:
The accepted values are:
For example to convert from BT.601 to SMPTE-240M, use the command:
colormatrix=bt601:smpte240m
Convert colorspace, transfer characteristics or color primaries. Input video needs to have an even size.
The filter accepts the following options:
The accepted values are:
The accepted values are:
The accepted values are:
The accepted values are:
The accepted values are:
The accepted values are:
The accepted values are:
The accepted values are:
The filter converts the transfer characteristics, color space and color primaries to the specified user values. The output value, if not specified, is set to a default value based on the "all" property. If that property is also not specified, the filter will log an error. The output color range and format default to the same value as the input color range and format. The input transfer characteristics, color space, color primaries and color range should be set on the input data. If any of these are missing, the filter will log an error and no conversion will take place.
For example to convert the input to SMPTE-240M, use the command:
colorspace=smpte240m
CUDA accelerated implementation of the colorspace filter.
It is by no means feature complete compared to the software colorspace filter, and at the current time only supports color range conversion between jpeg/full and mpeg/limited range.
The filter accepts the following options:
The accepted values are:
Adjust color temperature in video to simulate variations in ambient color temperature.
The filter accepts the following options:
Commands
This filter supports same commands as options.
Apply convolution of 3x3, 5x5, 7x7 or horizontal/vertical up to 49 elements.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Examples
convolution="0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0"
convolution="1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1/9:1/9:1/9:1/9"
convolution="0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:5:1:1:1:0:128:128:128"
convolution="0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:5:5:5:1:0:128:128:128"
convolution="1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:5:5:5:1:0:128:128:0"
convolution="-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2"
Apply 2D convolution of video stream in frequency domain using second stream as impulse.
The filter accepts the following options:
The "convolve" filter also supports the framesync options.
Copy the input video source unchanged to the output. This is mainly useful for testing purposes.
Video filtering on GPU using Apple's CoreImage API on OSX.
Hardware acceleration is based on an OpenGL context. Usually, this means it is processed by video hardware. However, software-based OpenGL implementations exist which means there is no guarantee for hardware processing. It depends on the respective OSX.
There are many filters and image generators provided by Apple that come with a large variety of options. The filter has to be referenced by its name along with its options.
The coreimage filter accepts the following options:
list_filters=true
It is required to specify either "default" or at least one of the filter options. All omitted options are used with their default values. The syntax of the filter string is as follows:
filter=<NAME>@<OPTION>=<VALUE>[@<OPTION>=<VALUE>][@...][#<NAME>@<OPTION>=<VALUE>[@<OPTION>=<VALUE>][@...]][#...]
output_rect=x\ y\ width\ height
If not given, the output rectangle equals the dimensions of the input image. The output rectangle is automatically cropped at the borders of the input image. Negative values are valid for each component.
output_rect=25\ 25\ 100\ 100
Several filters can be chained for successive processing without GPU-HOST transfers allowing for fast processing of complex filter chains. Currently, only filters with zero (generators) or exactly one (filters) input image and one output image are supported. Also, transition filters are not yet usable as intended.
Some filters generate output images with additional padding depending on the respective filter kernel. The padding is automatically removed to ensure the filter output has the same size as the input image.
For image generators, the size of the output image is determined by the previous output image of the filter chain or the input image of the whole filterchain, respectively. The generators do not use the pixel information of this image to generate their output. However, the generated output is blended onto this image, resulting in partial or complete coverage of the output image.
The coreimagesrc video source can be used for generating input images which are directly fed into the filter chain. By using it, providing input images by another video source or an input video is not required.
Examples
coreimage=list_filters=true
coreimage=filter=CIBoxBlur@default
coreimage=filter=CIBoxBlur@default#CIVignetteEffect@inputCenter=100\ 100@inputRadius=50
ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@inputMessage=https\\\\\://FFmpeg.org/@inputCorrectionLevel=H -frames:v 1 QRCode.png
Obtain the correlation between two input videos.
This filter takes two input videos.
Both input videos must have the same resolution and pixel format for this filter to work correctly. Also it assumes that both inputs have the same number of frames, which are compared one by one.
The obtained per component, average, min and max correlation is printed through the logging system.
The filter stores the calculated correlation of each frame in frame metadata.
This filter also supports the framesync options.
In the below example the input file main.mpg being processed is compared with the reference file ref.mpg.
ffmpeg -i main.mpg -i ref.mpg -lavfi corr -f null -
Cover a rectangular object
It accepts the following options:
It accepts the following values:
Default value is blur.
Examples
ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
Crop the input video to given dimensions.
It accepts the following parameters:
The out_w, out_h, x, y parameters are expressions containing the following constants:
The expression for out_w may depend on the value of out_h, and the expression for out_h may depend on out_w, but they cannot depend on x and y, as x and y are evaluated after out_w and out_h.
The x and y parameters specify the expressions for the position of the top-left corner of the output (non-cropped) area. They are evaluated for each frame. If the evaluated value is not valid, it is approximated to the nearest valid value.
The expression for x may depend on y, and the expression for y may depend on x.
Examples
crop=100:100:12:34
Using named options, the example above becomes:
crop=w=100:h=100:x=12:y=34
crop=100:100
crop=2/3*in_w:2/3*in_h
crop=out_w=in_h
crop=in_h
crop=in_w-100:in_h-100:100:100
crop=in_w-2*10:in_h-2*20
crop=in_w/2:in_h/2:in_w/2:in_h/2
crop=in_w:1/PHI*in_w
crop=in_w/2:in_h/2:(in_w-out_w)/2+((in_w-out_w)/2)*sin(n/10):(in_h-out_h)/2 +((in_h-out_h)/2)*sin(n/7)
crop=in_w/2:in_h/2:(in_w-out_w)/2+((in_w-out_w)/2)*sin(t*10):(in_h-out_h)/2 +((in_h-out_h)/2)*sin(t*13)
crop=in_w/2:in_h/2:y:10+10*sin(n/10)
Commands
This filter supports the following commands:
Auto-detect the crop size.
It calculates the necessary cropping parameters and prints the recommended parameters via the logging system. The detected dimensions correspond to the non-black or video area of the input video according to mode.
It accepts the following parameters:
This can be useful when channel logos distort the video area. 0 indicates 'never reset', and returns the largest area encountered during playback.
The high threshold selects the "strong" edge pixels, which are then connected through 8-connectivity with the "weak" edge pixels selected by the low threshold.
low and high threshold values must be chosen in the range [0,1], and low should be lesser or equal to high.
Default value for low is "5/255", and default value for high is "15/255".
Examples
ffmpeg -i file.mp4 -vf cropdetect,metadata=mode=print -f null -
ffmpeg -i file.mp4 -vf mestimate,cropdetect=mode=mvedges,metadata=mode=print -f null -
ffmpeg -flags2 +export_mvs -i file.mp4 -vf cropdetect=mode=mvedges,metadata=mode=print -f null -
Commands
This filter supports the following commands:
Delay video filtering until a given wallclock timestamp. The filter first passes on preroll amount of frames, then it buffers at most buffer amount of frames and waits for the cue. After reaching the cue it forwards the buffered frames and also any subsequent frames coming in its input.
The filter can be used synchronize the output of multiple ffmpeg processes for realtime output devices like decklink. By putting the delay in the filtering chain and pre-buffering frames the process can pass on data to output almost immediately after the target wallclock timestamp is reached.
Perfect frame accuracy cannot be guaranteed, but the result is good enough for some use cases.
Apply color adjustments using curves.
This filter is similar to the Adobe Photoshop and GIMP curves tools. Each component (red, green and blue) has its values defined by N key points tied from each other using a smooth curve. The x-axis represents the pixel values from the input frame, and the y-axis the new pixel values to be set for the output frame.
By default, a component curve is defined by the two points (0;0) and (1;1). This creates a straight line where each original pixel value is "adjusted" to its own value, which means no change to the image.
The filter allows you to redefine these two points and add some more. A new curve will be define to pass smoothly through all these new coordinates. The new defined points needs to be strictly increasing over the x-axis, and their x and y values must be in the [0;1] interval. The curve is formed by using a natural or monotonic cubic spline interpolation, depending on the interp option (default: "natural"). The "natural" spline produces a smoother curve in general while the monotonic ("pchip") spline guarantees the transitions between the specified points to be monotonic. If the computed curves happened to go outside the vector spaces, the values will be clipped accordingly.
The filter accepts the following options:
Default is "none".
To avoid some filtergraph syntax conflicts, each key points list need to be defined using the following syntax: "x0/y0 x1/y1 x2/y2 ...".
Commands
This filter supports same commands as options.
Examples
curves=blue='0/0 0.5/0.58 1/1'
curves=r='0/0.11 .42/.51 1/0.95':g='0/0 0.50/0.48 1/1':b='0/0.22 .49/.44 1/0.8'
Here we obtain the following coordinates for each components:
curves=preset=vintage
curves=vintage
curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
gnuplot -p /tmp/curves.plt
Video data analysis filter.
This filter shows hexadecimal pixel values of part of video.
The filter accepts the following options:
Commands
This filter supports same commands as options excluding "size" option.
Apply Directional blur filter.
The filter accepts the following options:
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Denoise frames using 2D DCT (frequency domain filtering).
This filter is not designed for real time.
The filter accepts the following options:
This sigma defines a hard threshold of "3 * sigma"; every DCT coefficient (absolute value) below this threshold with be dropped.
If you need a more advanced filtering, see expr.
Default is 0.
If the overlapping value doesn't permit processing the whole input width or height, a warning will be displayed and according borders won't be denoised.
Default value is blocksize-1, which is the best possible setting.
For each coefficient of a DCT block, this expression will be evaluated as a multiplier value for the coefficient.
If this is option is set, the sigma option will be ignored.
The absolute value of the coefficient can be accessed through the c variable.
The default value is 3 (8x8) and can be raised to 4 for a blocksize of 16x16. Note that changing this setting has huge consequences on the speed processing. Also, a larger block size does not necessarily means a better de-noising.
Examples
Apply a denoise with a sigma of 4.5:
dctdnoiz=4.5
The same operation can be achieved using the expression system:
dctdnoiz=e='gte(c, 4.5*3)'
Violent denoise using a block size of "16x16":
dctdnoiz=15:n=4
Remove banding artifacts from input video. It works by replacing banded pixels with average value of referenced pixels.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Remove blocking artifacts from input video.
The filter accepts the following options:
Examples
deblock=filter=weak:block=4
deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05
deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=1
deblock=filter=strong:block=4:alpha=0.12:beta=0.07:gamma=0.06:delta=0.05:planes=6
Commands
This filter supports the all above options as commands.
Drop duplicated frames at regular intervals.
The filter accepts the following options:
Apply 2D deconvolution of video stream in frequency domain using second stream as impulse.
The filter accepts the following options:
The "deconvolve" filter also supports the framesync options.
Reduce cross-luminance (dot-crawl) and cross-color (rainbows) from video.
It accepts the following options:
Apply deflate effect to the video.
This filter replaces the pixel by the local(3x3) average by taking into account only values lower than the pixel.
It accepts the following options:
Commands
This filter supports the all above options as commands.
Remove temporal frame luminance variations.
It accepts the following options:
Available values are:
Remove judder produced by partially interlaced telecined content.
Judder can be introduced, for instance, by pullup filter. If the original source was partially telecined content then the output of "pullup,dejudder" will have a variable frame rate. May change the recorded frame rate of the container. Aside from that change, this filter will not affect constant frame rate video.
The option available in this filter is:
Accepts any integer greater than 1. Useful values are:
The default is 4.
Suppress a TV station logo by a simple interpolation of the surrounding pixels. Just set a rectangle covering the logo and watch it disappear (and sometimes something even uglier appear - your mileage may vary).
It accepts the following parameters:
The rectangle is drawn on the outermost pixels which will be (partly) replaced with interpolated values. The values of the next pixels immediately outside this rectangle in each direction will be used to compute the interpolated pixel values inside the rectangle.
Examples
delogo=x=0:y=0:w=100:h=77
Remove the rain in the input image/video by applying the derain methods based on convolutional neural networks. Supported models:
Training as well as model generation scripts are provided in the repository at <https://github.com/XueweiMeng/derain_filter.git>.
The filter accepts the following options:
Default value is derain.
To get full functionality (such as async execution), please use the dnn_processing filter.
Attempt to fix small changes in horizontal and/or vertical shift. This filter helps remove camera shake from hand-holding a camera, bumping a tripod, moving on a vehicle, etc.
The filter accepts the following options:
This is useful when simultaneous movement of subjects within the frame might be confused for camera motion by the motion vector search.
If any or all of x, y, w and h are set to -1 then the full frame is used. This allows later options to be set without specifying the bounding box for the motion vector search.
Default - search the whole frame.
Default value is mirror.
Default value is exhaustive.
Remove unwanted contamination of foreground colors, caused by reflected color of greenscreen or bluescreen.
This filter accepts the following options:
Commands
This filter supports the all above options as commands.
Apply an exact inverse of the telecine operation. It requires a predefined pattern specified using the pattern option which must be the same as that passed to the telecine filter.
This filter accepts the following options:
Apply dilation effect to the video.
This filter replaces the pixel by the local(3x3) maximum.
It accepts the following options:
Flags to local 3x3 coordinates maps like this:
1 2 3
4 5
6 7 8
Commands
This filter supports the all above options as commands.
Displace pixels as indicated by second and third input stream.
It takes three input streams and outputs one stream, the first input is the source, and second and third input are displacement maps.
The second input specifies how much to displace pixels along the x-axis, while the third input specifies how much to displace pixels along the y-axis. If one of displacement map streams terminates, last frame from that displacement map will be used.
Note that once generated, displacements maps can be reused over and over again.
A description of the accepted options follows.
Available values are:
Default is smear.
Examples
ffmpeg -i INPUT -f lavfi -i nullsrc=s=hd720,lutrgb=128:128:128 -f lavfi -i nullsrc=s=hd720,geq='r=128+30*sin(2*PI*X/400+T):g=128+30*sin(2*PI*X/400+T):b=128+30*sin(2*PI*X/400+T)' -lavfi '[0][1][2]displace' OUTPUT
ffmpeg -i INPUT -f lavfi -i nullsrc=hd720,geq='r=128+80*(sin(sqrt((X-W/2)*(X-W/2)+(Y-H/2)*(Y-H/2))/220*2*PI+T)):g=128+80*(sin(sqrt((X-W/2)*(X-W/2)+(Y-H/2)*(Y-H/2))/220*2*PI+T)):b=128+80*(sin(sqrt((X-W/2)*(X-W/2)+(Y-H/2)*(Y-H/2))/220*2*PI+T))' -lavfi '[1]split[x][y],[0][x][y]displace' OUTPUT
Do classification with deep neural networks based on bounding boxes.
The filter accepts the following options:
For tensorflow backend, you can set its configs with sess_config options, please use tools/python/tf_sess_config.py to get the configs for your system.
Do object detection with deep neural networks.
The filter accepts the following options:
Do image processing with deep neural networks. It works together with another filter which converts the pixel format of the Frame to what the dnn network requires.
The filter accepts the following options:
For tensorflow backend, you can set its configs with sess_config options, please use tools/python/tf_sess_config.py to get the configs of TensorFlow backend for your system.
Examples
./ffmpeg -i rain.jpg -vf format=rgb24,dnn_processing=dnn_backend=tensorflow:model=can.pb:input=x:output=y derain.jpg
./ffmpeg -i 480p.jpg -vf format=yuv420p,scale=w=iw*2:h=ih*2,dnn_processing=dnn_backend=tensorflow:model=srcnn.pb:input=x:output=y -y srcnn.jpg
./ffmpeg -i 480p.jpg -vf format=yuv420p,dnn_processing=dnn_backend=tensorflow:model=espcn.pb:input=x:output=y:backend_configs=sess_config=0x10022805320e09cdccccccccccec3f20012a01303801 -y tmp.espcn.jpg
Draw a colored box on the input image.
It accepts the following parameters:
See below for the list of accepted constants.
The parameters for x, y, w and h and t are expressions containing the following constants:
If box_source is set, the x, y, width and height will be ignored and still use box data in detection bboxes of side data. So please do not use this parameter if you were not sure about the box source.
These constants allow the x, y, w, h and t expressions to refer to each other, so you may for example specify "y=x/dar" or "h=w/dar".
Examples
drawbox
drawbox=10:20:200:60:red@0.5
The previous example can be specified as:
drawbox=x=10:y=20:w=200:h=60:color=red@0.5
drawbox=x=10:y=10:w=100:h=100:color=pink@0.5:t=fill
drawbox=x=-t:y=0.5*(ih-iw/2.4)-t:w=iw+t*2:h=iw/2.4+t*2:t=2:c=red
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Draw a graph using input video metadata.
It accepts the following parameters:
Available values for mode is:
Default is "line".
Available values for slide is:
Default is "frame".
The foreground color expressions can use the following variables:
The color is defined as 0xAABBGGRR.
Example using metadata from signalstats filter:
signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
Example using metadata from ebur128 filter:
ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
Draw a grid on the input image.
It accepts the following parameters:
See below for the list of accepted constants.
The parameters for x, y, w and h and t are expressions containing the following constants:
These constants allow the x, y, w, h and t expressions to refer to each other, so you may for example specify "y=x/dar" or "h=w/dar".
Examples
drawgrid=width=100:height=100:thickness=2:color=red@0.5
drawgrid=w=iw/3:h=ih/3:t=2:c=white@0.5
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Draw a text string or text from a specified file on top of a video, using the libfreetype library.
To enable compilation of this filter, you need to configure FFmpeg with "--enable-libfreetype" and "--enable-libharfbuzz". To enable default font fallback and the font option you need to configure FFmpeg with "--enable-libfontconfig". To enable the text_shaping option, you need to configure FFmpeg with "--enable-libfribidi".
Syntax
It accepts the following parameters:
and the width of the left and right borders to 20
of the bottom border to 30 and the width of the left and right borders to 20
30 (bottom), 40 (left)
The default value of boxborderw is "0".
The default value of boxcolor is "white".
ascent (in pixels) defined in the font metrics
The default value of y_align is "text" for backward compatibility.
The default value of bordercolor is "black".
The default value of fontcolor is "black".
The flags map the corresponding flags supported by libfreetype, and are a combination of the following values:
Default value is "default".
For more information consult the documentation for the FT_LOAD_* libfreetype flags.
The default value of shadowcolor is "black".
This parameter is mandatory if no text string is specified with the parameter text.
If both text and textfile are specified, an error is thrown.
If text source is set, text and textfile will be ignored and still use text data in detection bboxes of side data. So please do not use this parameter if you are not sure about the text source.
The default value of x and y is "0".
See below for the list of accepted constants and functions.
The parameters for x and y are expressions containing the following constants and functions:
These parameters allow the x and y expressions to refer to each other, so you can for example specify "y=x/dar".
Text expansion
If expansion is set to "strftime", the filter recognizes sequences accepted by the "strftime" C function in the provided text and expands them accordingly. Check the documentation of "strftime". This feature is deprecated in favor of "normal" expansion with the "gmtime" or "localtime" expansion functions.
If expansion is set to "none", the text is printed verbatim.
If expansion is set to "normal" (which is the default), the following expansion mechanism is used.
The backslash character \, followed by any character, always expands to the second character.
Sequences of the form "%{...}" are expanded. The text between the braces is a function name, possibly followed by arguments separated by ':'. If the arguments contain special characters or delimiters (':' or '}'), they should be escaped.
Note that they probably must also be escaped as the value for the text option in the filter argument string and as the filter argument in the filtergraph description, and possibly also for the shell, that makes up to four levels of escaping; using a text file with the textfile option avoids these problems.
The following functions are available:
It must take one argument specifying the expression to be evaluated, which accepts the same constants and functions as the x and y values. Note that not all constants should be used, for example the text size is not known when evaluating the expression, so the constants text_w and text_h will have an undefined value.
The first argument is the expression to be evaluated, just as for the expr function. The second argument specifies the output format. Allowed values are x, X, d and u. They are treated exactly as in the "printf" function. The third parameter is optional and sets the number of positions taken by the output. It can be used to add padding with zeros from the left.
The first argument is mandatory and specifies the metadata key.
The second argument is optional and specifies a default value, used when the metadata key is not found or empty.
Available metadata can be identified by inspecting entries starting with TAG included within each frame section printed by running "ffprobe -show_frames".
String metadata generated in filters leading to the drawtext filter are also available.
The first argument is the format of the timestamp; it defaults to "flt" for seconds as a decimal number with microsecond accuracy; "hms" stands for a formatted [-]HH:MM:SS.mmm timestamp with millisecond accuracy. "gmtime" stands for the timestamp of the frame formatted as UTC time; "localtime" stands for the timestamp of the frame formatted as local time zone time.
The second argument is an offset added to the timestamp.
If the format is set to "hms", a third argument "24HH" may be supplied to present the hour part of the formatted timestamp in 24h format (00-23).
If the format is set to "localtime" or "gmtime", a third argument may be supplied: a "strftime" C function format string. By default, YYYY-MM-DD HH:MM:SS format will be used.
Commands
This filter supports altering parameters via commands:
Syntax for the argument is the same as for filter invocation, e.g.
fontsize=56:fontcolor=green:text='Hello World'
Full filter invocation with sendcmd would look like this:
sendcmd=c='56.0 drawtext reinit fontsize=56\:fontcolor=green\:text=Hello\\ World'
If the entire argument can't be parsed or applied as valid values then the filter will continue with its existing parameters.
The following options are also supported as commands:
Examples
drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
x=100: y=50: fontsize=24: fontcolor=yellow@0.2: box=1: boxcolor=red@0.2"
Note that the double quotes are not necessary if spaces are not used within the parameter list.
drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=if(eq(mod(t\,30)\,0)\,rand(0\,(w-text_w))\,x):y=if(eq(mod(t\,30)\,0)\,rand(0\,(h-text_h))\,y)"
drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
drawtext="text='Test Text': fontsize=h/30: x=(w-text_w)/2: y=(h-text_h*2)"
drawtext='fontfile=FreeSans.ttf:text=%{localtime\:%a %b %d %Y}'
#!/bin/sh
DS=1.0 # display start
DE=10.0 # display end
FID=1.5 # fade in duration
FOD=5 # fade out duration
ffplay -f lavfi "color,drawtext=text=TEST:fontsize=50:fontfile=FreeSerif.ttf:fontcolor_expr=ff0000%{eif\\\\: clip(255*(1*between(t\\, $DS + $FID\\, $DE - $FOD) + ((t - $DS)/$FID)*between(t\\, $DS\\, $DS + $FID) + (-(t - $DE)/$FOD)*between(t\\, $DE - $FOD\\, $DE) )\\, 0\\, 255) \\\\: x\\\\: 2 }"
drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
drawtext="fontsize=20:fontcolor=white:fontfile=FreeSans.ttf:text='%{metadata\:lavf.image2dec.source_basename\:NA}':x=10:y=10"
For more information about libfreetype, check: <http://www.freetype.org/>.
For more information about fontconfig, check: <http://freedesktop.org/software/fontconfig/fontconfig-user.html>.
For more information about libfribidi, check: <http://fribidi.org/>.
For more information about libharfbuzz, check: <https://github.com/harfbuzz/harfbuzz>.
Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
The filter accepts the following options:
The high threshold selects the "strong" edge pixels, which are then connected through 8-connectivity with the "weak" edge pixels selected by the low threshold.
low and high threshold values must be chosen in the range [0,1], and low should be lesser or equal to high.
Default value for low is "20/255", and default value for high is "50/255".
Default value is wires.
Examples
edgedetect=low=0.1:high=0.4
edgedetect=mode=colormix:high=0
Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
For each input image, the filter will compute the optimal mapping from the input to the output given the codebook length, that is the number of distinct output colors.
This filter accepts the following options.
Measure graylevel entropy in histogram of color channels of video frames.
It accepts the following parameters:
diff mode measures entropy of histogram delta values, absolute differences between neighbour histogram values.
Apply the EPX magnification filter which is designed for pixel art.
It accepts the following option:
Set brightness, contrast, saturation and approximate gamma adjustment.
The filter accepts the following options:
It accepts the following values:
Default value is init.
The expressions accept the following parameters:
Commands
The filter supports the following commands:
The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Apply erosion effect to the video.
This filter replaces the pixel by the local(3x3) minimum.
It accepts the following options:
Flags to local 3x3 coordinates maps like this:
1 2 3
4 5
6 7 8
Commands
This filter supports the all above options as commands.
Deinterlace the input video ("estdif" stands for "Edge Slope Tracing Deinterlacing Filter").
Spatial only filter that uses edge slope tracing algorithm to interpolate missing lines. It accepts the following parameters:
The default value is "field".
The default value is "auto". If the interlacing is unknown or the decoder does not export this information, top field first will be assumed.
The default value is "all".
Commands
This filter supports same commands as options.
Adjust exposure of the video stream.
The filter accepts the following options:
Commands
This filter supports same commands as options.
Extract color channel components from input video stream into separate grayscale video streams.
The filter accepts the following option:
Available values for planes are:
Choosing planes not available in the input will result in an error. That means you cannot select "r", "g", "b" planes with "y", "u", "v" planes at same time.
Examples
ffmpeg -i video.avi -filter_complex 'extractplanes=y+u+v[y][u][v]' -map '[y]' y.avi -map '[u]' u.avi -map '[v]' v.avi
Apply a fade-in/out effect to the input video.
It accepts the following parameters:
Examples
fade=in:0:30
The command above is equivalent to:
fade=t=in:s=0:n=30
fade=out:155:45
fade=type=out:start_frame=155:nb_frames=45
fade=in:0:25, fade=out:975:25
fade=in:5:20:color=yellow
fade=in:0:25:alpha=1
fade=t=in:st=5.5:d=0.5
Apply feedback video filter.
This filter pass cropped input frames to 2nd output. From there it can be filtered with other video filters. After filter receives frame from 2nd input, that frame is combined on top of original frame from 1st input and passed to 1st output.
The typical usage is filter only part of frame.
The filter accepts the following options:
Examples
[in][blurin]feedback=x=0:y=0:w=100:h=100[out][blurout];[blurout]gblur=8[blurin]
[in][blurin]feedback=x=0:y=0:w=100:h=100[out][blurout];[blurout]drawbox=x=0:y=0:w=100:h=100:t=100[blurin]
Denoise frames using 3D FFT (frequency domain filtering).
The filter accepts the following options:
Apply arbitrary expressions to samples in frequency domain
It accepts the following values:
Default value is init.
The filter accepts the following variables:
Examples
fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
Extract a single field from an interlaced image using stride arithmetic to avoid wasting CPU time. The output frames are marked as non-interlaced.
The filter accepts the following options:
Create new frames by copying the top and bottom fields from surrounding frames supplied as numbers by the hint file.
There must be one line for each frame in a clip. Each line must contain two numbers separated by the comma, optionally followed by "-" or "+". Numbers supplied on each line of file can not be out of [N-1,N+1] where N is current frame number for "absolute" mode or out of [-1, 1] range for "relative" mode. First number tells from which frame to pick up top field and second number tells from which frame to pick up bottom field.
If optionally followed by "+" output frame will be marked as interlaced, else if followed by "-" output frame will be marked as progressive, else it will be marked same as input frame. If optionally followed by "t" output frame will use only top field, or in case of "b" it will use only bottom field. If line starts with "#" or ";" that line is skipped.
Example of first several lines of "hint" file for "relative" mode:
0,0 - # first frame
1,0 - # second frame, use third's frame top field and second's frame bottom field
1,0 - # third frame, use fourth's frame top field and third's frame bottom field
1,0 -
0,0 -
0,0 -
1,0 -
1,0 -
1,0 -
0,0 -
0,0 -
1,0 -
1,0 -
1,0 -
0,0 -
Field matching filter for inverse telecine. It is meant to reconstruct the progressive frames from a telecined stream. The filter does not drop duplicated frames, so to achieve a complete inverse telecine "fieldmatch" needs to be followed by a decimation filter such as decimate in the filtergraph.
The separation of the field matching and the decimation is notably motivated by the possibility of inserting a de-interlacing filter fallback between the two. If the source has mixed telecined and real interlaced content, "fieldmatch" will not be able to match fields for the interlaced parts. But these remaining combed frames will be marked as interlaced, and thus can be de-interlaced by a later filter such as yadif before decimation.
In addition to the various configuration options, "fieldmatch" can take an optional second stream, activated through the ppsrc option. If enabled, the frames reconstruction will be based on the fields and frames from this second stream. This allows the first input to be pre-processed in order to help the various algorithms of the filter, while keeping the output lossless (assuming the fields are matched properly). Typically, a field-aware denoiser, or brightness/contrast adjustments can help.
Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project) and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from which "fieldmatch" is based on. While the semantic and usage are very close, some behaviour and options names can differ.
The decimate filter currently only works for constant frame rate input. If your input has mixed telecined (30fps) and progressive content with a lower framerate like 24fps use the following filterchain to produce the necessary cfr stream: "dejudder,fps=30000/1001,fieldmatch,decimate".
The filter accepts the following options:
Note that it is sometimes recommended not to trust the parity announced by the stream.
Default value is auto.
More details about p/c/n/u/b are available in p/c/n/u/b meaning section.
Available values are:
The parenthesis at the end indicate the matches that would be used for that mode assuming order=tff (and field on auto or top).
In terms of speed pc mode is by far the fastest and pcn_ub is the slowest.
Default value is pc_n.
Default value is 0 (disabled).
Default value is auto.
Default value is 1.
Default value is 12.0.
Default is sc.
Default value is none.
Default value is 9.
Default value is 0.
Default value is 16.
Default value is 80.
p/c/n/u/b meaning
p/c/n
We assume the following telecined stream:
Top fields: 1 2 2 3 4
Bottom fields: 1 2 3 4 4
The numbers correspond to the progressive frame the fields relate to. Here, the first two frames are progressive, the 3rd and 4th are combed, and so on.
When "fieldmatch" is configured to run a matching from bottom (field=bottom) this is how this input stream get transformed:
Input stream:
T 1 2 2 3 4
B 1 2 3 4 4 <-- matching reference
Matches: c c n n c
Output stream:
T 1 2 3 4 4
B 1 2 3 4 4
As a result of the field matching, we can see that some frames get duplicated. To perform a complete inverse telecine, you need to rely on a decimation filter after this operation. See for instance the decimate filter.
The same operation now matching from top fields (field=top) looks like this:
Input stream:
T 1 2 2 3 4 <-- matching reference
B 1 2 3 4 4
Matches: c c p p c
Output stream:
T 1 2 2 3 4
B 1 2 2 3 4
In these examples, we can see what p, c and n mean; basically, they refer to the frame and field of the opposite parity:
u/b
The u and b matching are a bit special in the sense that they match from the opposite parity flag. In the following examples, we assume that we are currently matching the 2nd frame (Top:2, bottom:2). According to the match, a 'x' is placed above and below each matched fields.
With bottom matching (field=bottom):
Match: c p n b u
x x x x x
Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
x x x x x
Output frames:
2 1 2 2 2
2 2 2 1 3
With top matching (field=top):
Match: c p n b u
x x x x x
Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
x x x x x
Output frames:
2 2 2 1 2
2 1 3 2 2
Examples
Simple IVTC of a top field first telecined stream:
fieldmatch=order=tff:combmatch=none, decimate
Advanced IVTC, with fallback on yadif for still combed frames:
fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
Transform the field order of the input video.
It accepts the following parameters:
The default value is tff.
The transformation is done by shifting the picture content up or down by one line, and filling the remaining line with appropriate picture content. This method is consistent with most broadcast field order converters.
If the input video is not flagged as being interlaced, or it is already flagged as being of the required output field order, then this filter does not alter the incoming video.
It is very useful when converting to or from PAL DV material, which is bottom field first.
For example:
ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
Buffer input images and send them when they are requested.
It is mainly useful when auto-inserted by the libavfilter framework.
It does not take parameters.
Fill borders of the input video, without changing video stream dimensions. Sometimes video can have garbage at the four edges and you may not want to crop video input to keep size multiple of some number.
This filter accepts the following options:
It accepts the following values:
Default is smear.
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Find a rectangular object in the input video.
The object to search for must be specified as a gray8 image specified with the object option.
For each possible match, a score is computed. If the score reaches the specified threshold, the object is considered found.
If the input video contains multiple instances of the object, the filter will find only one of them.
When an object is found, the following metadata entries are set in the matching frame:
It accepts the following options:
A threshold value of 0.01 means only exact matches, a threshold of 0.99 means almost everything matches.
Default value is 0.5.
Examples
ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
ffprobe -f lavfi movie=test.mp4,find_rect=object=object.pgm:threshold=0.3 \
-show_entries frame=pkt_pts_time:frame_tags=lavfi.rect.x,lavfi.rect.y \
-of csv -o find_rect.csv
Flood area with values of same pixel components with another values.
It accepts the following options:
Convert the input video to one of the specified pixel formats. Libavfilter will try to pick one that is suitable as input to the next filter.
It accepts the following parameters:
Examples
format=pix_fmts=yuv420p
Convert the input video to any of the formats in the list
format=pix_fmts=yuv420p|yuv444p|yuv410p
Convert the video to specified constant frame rate by duplicating or dropping frames as necessary.
It accepts the following parameters:
The default is 25.
Possible values are:
The default is "near".
Possible values are:
The default is "round".
Alternatively, the options can be specified as a flat string: fps[:start_time[:round]].
See also the setpts filter.
Examples
fps=fps=25
fps=fps=film:round=near
Pack two different video streams into a stereoscopic video, setting proper metadata on supported codecs. The two views should have the same size and framerate and processing will stop when the shorter video ends. Please note that you may conveniently adjust view properties with the scale and fps filters.
It accepts the following parameters:
Some examples:
# Convert left and right views into a frame-sequential video
ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
# Convert views into a side-by-side video with the same output resolution as the input
ffmpeg -i LEFT -i RIGHT -filter_complex [0:v]scale=w=iw/2[left],[1:v]scale=w=iw/2[right],[left][right]framepack=sbs OUTPUT
Change the frame rate by interpolating new video output frames from the source frames.
This filter is not designed to function correctly with interlaced media. If you wish to change the frame rate of interlaced media then you are required to deinterlace before this filter and re-interlace after this filter.
A description of the accepted options follows.
Available value for flags is:
Select one frame every N-th frame.
This filter accepts the following option:
Detect frozen video.
This filter logs a message and sets frame metadata when it detects that the input video has no significant change in content during a specified duration. Video freeze detection calculates the mean average absolute difference of all the components of video frames and compares it to a noise floor.
The printed times and duration are expressed in seconds. The "lavfi.freezedetect.freeze_start" metadata key is set on the first frame whose timestamp equals or exceeds the detection duration and it contains the timestamp of the first frame of the freeze. The "lavfi.freezedetect.freeze_duration" and "lavfi.freezedetect.freeze_end" metadata keys are set on the first frame after the freeze.
The filter accepts the following options:
Freeze video frames.
This filter freezes video frames using frame from 2nd input.
The filter accepts the following options:
Apply a frei0r effect to the input video.
To enable the compilation of this filter, you need to install the frei0r header and configure FFmpeg with "--enable-frei0r".
It accepts the following parameters:
A frei0r effect parameter can be a boolean (its value is either "y" or "n"), a double, a color (specified as R/G/B, where R, G, and B are floating point numbers between 0.0 and 1.0, inclusive) or a color description as specified in the "Color" section in the ffmpeg-utils manual, a position (specified as X/Y, where X and Y are floating point numbers) and/or a string.
The number and types of parameters depend on the loaded effect. If an effect parameter is not specified, the default value is set.
Examples
frei0r=filter_name=distort0r:filter_params=0.5|0.01
frei0r=colordistance:0.2/0.3/0.4
frei0r=colordistance:violet
frei0r=colordistance:0x112233
frei0r=perspective:0.2/0.2|0.8/0.2
For more information, see <http://frei0r.dyne.org>
Commands
This filter supports the filter_params option as commands.
Apply fast and simple postprocessing. It is a faster version of spp.
It splits (I)DCT into horizontal/vertical passes. Unlike the simple post- processing filter, one of them is performed once per block, not per pixel. This allows for much higher speed.
The filter accepts the following options:
Apply Gaussian blur filter.
The filter accepts the following options:
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Apply generic equation to each pixel.
The filter accepts the following options:
The colorspace is selected according to the specified options. If one of the lum_expr, cb_expr, or cr_expr options is specified, the filter will automatically select a YCbCr colorspace. If one of the red_expr, green_expr, or blue_expr options is specified, it will select an RGB colorspace.
If one of the chrominance expression is not defined, it falls back on the other one. If no alpha expression is specified it will evaluate to opaque value. If none of chrominance expressions are specified, they will evaluate to the luma expression.
The expressions can use the following variables and functions:
Default is bilinear.
For functions, if x and y are outside the area, the value will be automatically clipped to the closer edge.
Please note that this filter can use multiple threads in which case each slice will have its own expression state. If you want to use only a single expression state because your expressions depend on previous state then you should limit the number of filter threads to 1.
Examples
geq=p(W-X\,Y)
geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
nullsrc=s=256x256,geq=random(1)/hypot(X-cos(N*0.07)*W/2-W/2\,Y-sin(N*0.09)*H/2-H/2)^2*1000000*sin(N*0.02):128:128
format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
Fix the banding artifacts that are sometimes introduced into nearly flat regions by truncation to 8-bit color depth. Interpolate the gradients that should go where the bands are, and dither them.
It is designed for playback only. Do not use it prior to lossy compression, because compression tends to lose the dither and bring back the bands.
It accepts the following parameters:
Alternatively, the options can be specified as a flat string: strength[:radius]
Examples
gradfun=3.5:8
gradfun=radius=8
Show various filtergraph stats.
With this filter one can debug complete filtergraph. Especially issues with links filling with queued frames.
The filter accepts the following options:
Available values for flags are:
Available values for flags are:
A color constancy filter that applies color correction based on the grayworld assumption
See: <https://www.researchgate.net/publication/275213614_A_New_Color_Correction_Method_for_Underwater_Imaging>
The algorithm uses linear light, so input data should be linearized beforehand (and possibly correctly tagged).
ffmpeg -i INPUT -vf zscale=transfer=linear,grayworld,zscale=transfer=bt709,format=yuv420p OUTPUT
A color constancy variation filter which estimates scene illumination via grey edge algorithm and corrects the scene colors accordingly.
See: <https://staff.science.uva.nl/th.gevers/pub/GeversTIP07.pdf>
The filter accepts the following options:
Examples
greyedge=difford=1:minknorm=5:sigma=2
greyedge=difford=1:minknorm=0:sigma=2
Apply guided filter for edge-preserving smoothing, dehazing and so on.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Examples
ffmpeg -i in.png -vf guided out.png
ffmpeg -i in.png -i guidance.png -filter_complex guided=guidance=on out.png
Apply a Hald CLUT to a video stream.
First input is the video stream to process, and second one is the Hald CLUT. The Hald CLUT input can be a simple picture or a complete video stream.
The filter accepts the following options:
"haldclut" also has the same interpolation options as lut3d (both filters share the same internals).
This filter also supports the framesync options.
More information about the Hald CLUT can be found on Eskil Steenberg's website (Hald CLUT author) at <http://www.quelsolaar.com/technology/clut.html>.
Commands
This filter supports the "interp" option as commands.
Workflow examples
Hald CLUT video stream
Generate an identity Hald CLUT stream altered with various effects:
ffmpeg -f lavfi -i B<haldclutsrc>=8 -vf "hue=H=2*PI*t:s=sin(2*PI*t)+1, curves=cross_process" -t 10 -c:v ffv1 clut.nut
Note: make sure you use a lossless codec.
Then use it with "haldclut" to apply it on some random stream:
ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
The Hald CLUT will be applied to the 10 first seconds (duration of clut.nut), then the latest picture of that CLUT stream will be applied to the remaining frames of the "mandelbrot" stream.
Hald CLUT with preview
A Hald CLUT is supposed to be a squared image of "Level*Level*Level" by "Level*Level*Level" pixels. For a given Hald CLUT, FFmpeg will select the biggest possible square starting at the top left of the picture. The remaining padding pixels (bottom or right) will be ignored. This area can be used to add a preview of the Hald CLUT.
Typically, the following generated Hald CLUT will be supported by the "haldclut" filter:
ffmpeg -f lavfi -i B<haldclutsrc>=8 -vf "
pad=iw+320 [padded_clut];
smptebars=s=320x256, split [a][b];
[padded_clut][a] overlay=W-320:h, curves=color_negative [main];
[main][b] overlay=W-320" -frames:v 1 clut.png
It contains the original and a preview of the effect of the CLUT: SMPTE color bars are displayed on the right-top, and below the same color bars processed by the color changes.
Then, the effect of this Hald CLUT can be visualized with:
ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
Flip the input video horizontally.
For example, to horizontally flip the input video with ffmpeg:
ffmpeg -i in.avi -vf "hflip" out.avi
This filter applies a global color histogram equalization on a per-frame basis.
It can be used to correct video that has a compressed range of pixel intensities. The filter redistributes the pixel intensities to equalize their distribution across the intensity range. It may be viewed as an "automatically adjusting contrast filter". This filter is useful only for correcting degraded or poorly captured source video.
The filter accepts the following options:
Compute and draw a color distribution histogram for the input video.
The computed histogram is a representation of the color component distribution in an image.
Standard histogram displays the color components distribution in an image. Displays color graph for each color component. Shows distribution of the Y, U, V, A or R, G, B components, depending on input format, in the current frame. Below each graph a color component scale meter is shown.
The filter accepts the following options:
Default is "stack".
Default is "whiteonblack".
Examples
ffplay -i input -vf histogram
This is a high precision/quality 3d denoise filter. It aims to reduce image noise, producing smooth images and making still images really still. It should enhance compressibility.
It accepts the following optional parameters:
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Download hardware frames to system memory.
The input must be in hardware frames, and the output a non-hardware format. Not all formats will be supported on the output - it may be necessary to insert an additional format filter immediately following in the graph to get the output in a supported format.
Map hardware frames to system memory or to another device.
This filter has several different modes of operation; which one is used depends on the input and output formats:
Map the input frames to system memory and pass them to the output. If the original hardware frame is later required (for example, after overlaying something else on part of it), the hwmap filter can be used again in the next mode to retrieve it.
If the input is actually a software-mapped hardware frame, then unmap it - that is, return the original hardware frame.
Otherwise, a device must be provided. Create new hardware surfaces on that device for the output, then map them back to the software format at the input and give those frames to the preceding filter. This will then act like the hwupload filter, but may be able to avoid an additional copy when the input is already in a compatible format.
A device must be supplied for the output, either directly or with the derive_device option. The input and output devices must be of different types and compatible - the exact meaning of this is system-dependent, but typically it means that they must refer to the same underlying hardware context (for example, refer to the same graphics card).
If the input frames were originally created on the output device, then unmap to retrieve the original frames.
Otherwise, map the frames to the output device - create new hardware frames on the output corresponding to the frames on the input.
The following additional parameters are accepted:
This may improve performance in some cases, as the original contents of the frame need not be loaded.
Indirect mappings to copies of frames are created in some cases where either direct mapping is not possible or it would have unexpected properties. Setting this flag ensures that the mapping is direct and will fail if that is not possible.
Defaults to read+write if not specified.
This option is dangerous - it may break the preceding filter in undefined ways if there are any additional constraints on that filter's output. Do not use it without fully understanding the implications of its use.
Upload system memory frames to hardware surfaces.
The device to upload to must be supplied when the filter is initialised. If using ffmpeg, select the appropriate device with the -filter_hw_device option or with the derive_device option. The input and output devices must be of different types and compatible - the exact meaning of this is system-dependent, but typically it means that they must refer to the same underlying hardware context (for example, refer to the same graphics card).
The following additional parameters are accepted:
Upload system memory frames to a CUDA device.
It accepts the following optional parameters:
Apply a high-quality magnification filter designed for pixel art. This filter was originally created by Maxim Stepin.
It accepts the following option:
Stack input videos horizontally.
All streams must be of same pixel format and of same height.
Note that this filter is faster than using overlay and pad filter to create same output.
The filter accepts the following option:
Turns a certain HSV range into gray values.
This filter measures color difference between set HSV color in options and ones measured in video stream. Depending on options, output colors can be changed to be gray or not.
The filter accepts the following options:
0.00001 matches only the exact key color, while 1.0 matches everything.
0.0 makes pixels either fully gray, or not gray at all.
Higher values result in more gray pixels, with a higher gray pixel the more similar the pixels color is to the key color.
Turns a certain HSV range into transparency.
This filter measures color difference between set HSV color in options and ones measured in video stream. Depending on options, output colors can be changed to transparent by adding alpha channel.
The filter accepts the following options:
0.00001 matches only the exact key color, while 1.0 matches everything.
0.0 makes pixels either fully transparent, or not transparent at all.
Higher values result in semi-transparent pixels, with a higher transparency the more similar the pixels color is to the key color.
Modify the hue and/or the saturation of the input.
It accepts the following parameters:
h and H are mutually exclusive, and can't be specified at the same time.
The b, h, H and s option values are expressions containing the following constants:
Examples
hue=h=90:s=1
hue=H=PI/2:s=1
hue="H=2*PI*t: s=sin(2*PI*t)+1"
hue="s=min(t/3\,1)"
The general fade-in expression can be written as:
hue="s=min(0\, max((t-START)/DURATION\, 1))"
hue="s=max(0\, min(1\, (8-t)/3))"
The general fade-out expression can be written as:
hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
Commands
This filter supports the following commands:
Apply hue-saturation-intensity adjustments to input video stream.
This filter operates in RGB colorspace.
This filter accepts the following options:
Grow first stream into second stream by connecting components. This makes it possible to build more robust edge masks.
This filter accepts the following options:
The "hysteresis" filter also supports the framesync options.
Detect the colorspace from an embedded ICC profile (if present), and update the frame's tags accordingly.
This filter accepts the following options:
Generate ICC profiles and attach them to frames.
This filter accepts the following options:
See the setparams filter for a list of possible values, but note that "unknown" are not valid values for this filter.
Obtain the identity score between two input videos.
This filter takes two input videos.
Both input videos must have the same resolution and pixel format for this filter to work correctly. Also it assumes that both inputs have the same number of frames, which are compared one by one.
The obtained per component, average, min and max identity score is printed through the logging system.
The filter stores the calculated identity scores of each frame in frame metadata.
This filter also supports the framesync options.
In the below example the input file main.mpg being processed is compared with the reference file ref.mpg.
ffmpeg -i main.mpg -i ref.mpg -lavfi identity -f null -
Detect video interlacing type.
This filter tries to detect if the input frames are interlaced, progressive, top or bottom field first. It will also try to detect fields that are repeated between adjacent frames (a sign of telecine).
Single frame detection considers only immediately adjacent frames when classifying each frame. Multiple frame detection incorporates the classification history of previous frames.
The filter will log these metadata values:
The filter accepts the following options:
Deinterleave or interleave fields.
This filter allows one to process interlaced images fields without deinterlacing them. Deinterleaving splits the input frame into 2 fields (so called half pictures). Odd lines are moved to the top half of the output image, even lines to the bottom half. You can process (filter) them independently and then re-interleave them.
The filter accepts the following options:
Default value is "none".
Commands
This filter supports the all above options as commands.
Apply inflate effect to the video.
This filter replaces the pixel by the local(3x3) average by taking into account only values higher than the pixel.
It accepts the following options:
Commands
This filter supports the all above options as commands.
Simple interlacing filter from progressive contents. This interleaves upper (or lower) lines from odd frames with lower (or upper) lines from even frames, halving the frame rate and preserving image height.
Original Original New Frame
Frame 'j' Frame 'j+1' (tff)
========== =========== ==================
Line 0 --------------------> Frame 'j' Line 0
Line 1 Line 1 ----> Frame 'j+1' Line 1
Line 2 ---------------------> Frame 'j' Line 2
Line 3 Line 3 ----> Frame 'j+1' Line 3
... ... ...
New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
It accepts the following optional parameters:
Deinterlace input video by applying Donald Graft's adaptive kernel deinterling. Work on interlaced parts of a video to produce progressive frames.
The description of the accepted parameters follows.
Examples
kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
kerndeint=sharp=1
kerndeint=map=1
Apply kirsch operator to input video stream.
The filter accepts the following option:
Commands
This filter supports the all above options as commands.
Slowly update darker pixels.
This filter makes short flashes of light appear longer. This filter accepts the following options:
Commands
This filter supports the all above options as commands.
Correct radial lens distortion
This filter can be used to correct for radial distortion as can result from the use of wide angle lenses, and thereby re-rectify the image. To find the right parameters one can use tools available for example as part of opencv or simply trial-and-error. To use opencv use the calibration sample (under samples/cpp) from the opencv sources and extract the k1 and k2 coefficients from the resulting matrix.
Note that effectively the same filter is available in the open-source tools Krita and Digikam from the KDE project.
In contrast to the vignette filter, which can also be used to compensate lens errors, this filter corrects the distortion of the image, whereas vignette corrects the brightness distribution, so you may want to use both filters together in certain cases, though you will have to take care of ordering, i.e. whether vignetting should be applied before or after lens correction.
Options
The filter accepts the following options:
The formula that generates the correction is:
r_src = r_tgt * (1 + k1 * (r_tgt / r_0)^2 + k2 * (r_tgt / r_0)^4)
where r_0 is halve of the image diagonal and r_src and r_tgt are the distances from the focal point in the source and target images, respectively.
Commands
This filter supports the all above options as commands.
Apply lens correction via the lensfun library (<http://lensfun.sourceforge.net/>).
The "lensfun" filter requires the camera make, camera model, and lens model to apply the lens correction. The filter will load the lensfun database and query it to find the corresponding camera and lens entries in the database. As long as these entries can be found with the given options, the filter can perform corrections on frames. Note that incomplete strings will result in the filter choosing the best match with the given options, and the filter will output the chosen camera and lens models (logged with level "info"). You must provide the make, camera model, and lens model as they are required.
To obtain a list of available makes and models, leave out one or both of "make" and "model" options. The filter will send the full list to the log with level "INFO". The first column is the make and the second column is the model. To obtain a list of available lenses, set any values for make and model and leave out the "lens_model" option. The filter will send the full list of lenses in the log with level "INFO". The ffmpeg tool will exit after the list is printed.
The filter accepts the following options:
Examples
ffmpeg -i input.mov -vf lensfun=make=Canon:model="Canon EOS 100D":lens_model="Canon EF-S 18-55mm f/3.5-5.6 IS STM":focal_length=18:aperture=8 -c:v h264 -b:v 8000k output.mov
ffmpeg -i input.mov -vf lensfun=make=Canon:model="Canon EOS 100D":lens_model="Canon EF-S 18-55mm f/3.5-5.6 IS STM":focal_length=18:aperture=8:enable='lte(t\,5)' -c:v h264 -b:v 8000k output.mov
Flexible GPU-accelerated processing filter based on libplacebo (<https://code.videolan.org/videolan/libplacebo>).
Options
The options for this filter are divided into the following sections:
Output mode
These options control the overall output mode. By default, libplacebo will try to preserve the source colorimetry and size as best as it can, but it will apply any embedded film grain, dolby vision metadata or anamorphic SAR present in source frames.
Allows for the same expressions as the scale filter.
-vf "libplacebo=w=iw*2:h=ih*2:extra_opts='upscaler=custom\:upscaler_preset=ewa_lanczos\:upscaler_blur=0.9812505644269356'"
See the setparams filter for a list of possible values.
In addition to the expression constants documented for the scale filter, the crop_w, crop_h, crop_x, crop_y, pos_w, pos_h, pos_x and pos_y options can also contain the following constants:
Scaling
The options in this section control how libplacebo performs upscaling and (if necessary) downscaling. Note that libplacebo will always internally operate on 4:4:4 content, so any sub-sampled chroma formats such as "yuv420p" will necessarily be upsampled and downsampled as part of the rendering process. That means scaling might be in effect even if the source and destination resolution are the same.
Debanding
Libplacebo comes with a built-in debanding filter that is good at counteracting many common sources of banding and blocking. Turning this on is highly recommended whenever quality is desired.
Color adjustment
A collection of subjective color controls. Not very rigorous, so the exact effect will vary somewhat depending on the input primaries and colorspace.
Peak detection
To help deal with sources that only have static HDR10 metadata (or no tagging whatsoever), libplacebo uses its own internal frame analysis compute shader to analyze source frames and adapt the tone mapping function in realtime. If this is too slow, or if exactly reproducible frame-perfect results are needed, it's recommended to turn this feature off.
Tone mapping
The options in this section control how libplacebo performs tone-mapping and gamut-mapping when dealing with mismatches between wide-gamut or HDR content. In general, libplacebo relies on accurate source tagging and mastering display gamut information to produce the best results.
Dithering
By default, libplacebo will dither whenever necessary, which includes rendering to any integer format below 16-bit precision. It's recommended to always leave this on, since not doing so may result in visible banding in the output, even if the "debanding" filter is enabled. If maximum performance is needed, use "ordered_fixed" instead of disabling dithering.
Custom shaders
libplacebo supports a number of custom shaders based on the mpv .hook GLSL syntax. A collection of such shaders can be found here: <https://github.com/mpv-player/mpv/wiki/User-Scripts#user-shaders>
A full description of the mpv shader format is beyond the scope of this section, but a summary can be found here: <https://mpv.io/manual/master/#options-glsl-shader>
Debugging / performance
All of the options in this section default off. They may be of assistance when attempting to squeeze the maximum performance at the cost of quality.
Commands
This filter supports almost all of the above options as commands.
Examples
libplacebo=colorspace=bt709:color_primaries=bt709:color_trc=bt709:range=tv
libplacebo=w=1920:h=1080:force_original_aspect_ratio=decrease:normalize_sar=true:upscaler=ewa_lanczos:downscaler=ewa_lanczos
libplacebo=fps=60:frame_mixer=mitchell_clamp
libplacebo=format=yuv420p:colorspace=bt470bg:color_primaries=bt709:color_trc=iec61966-2-1:range=pc
libplacebo=deband=true:deband_iterations=3:deband_radius=8:deband_threshold=6
ffmpeg ... -init_hw_device vulkan:llvmpipe ... -vf libplacebo=upscaler=none:downscaler=none:peak_detect=false
ffmpeg -export_side_data +film_grain ... -vf libplacebo=apply_filmgrain=true
ffmpeg -init_hw_device vulkan -hwaccel vaapi -hwaccel_output_format vaapi ... -vf libplacebo
Calulate the VMAF (Video Multi-Method Assessment Fusion) score for a reference/distorted pair of input videos.
The first input is the distorted video, and the second input is the reference video.
The obtained VMAF score is printed through the logging system.
It requires Netflix's vmaf library (libvmaf) as a pre-requisite. After installing the library it can be enabled using: "./configure --enable-libvmaf".
The filter has following options:
This filter also supports the framesync options.
Examples
ffmpeg -i distorted.mpg -i reference.mpg -lavfi libvmaf=log_path=output.xml -f null -
ffmpeg -i distorted.mpg -i reference.mpg -lavfi libvmaf='model=version=vmaf_v0.6.1\\:name=vmaf|version=vmaf_v0.6.1neg\\:name=vmaf_neg' -f null -
ffmpeg -i distorted.mpg -i reference.mpg -lavfi libvmaf='feature=name=psnr|name=ciede' -f null -
ffmpeg -i distorted.mpg -i reference.mkv -lavfi "[0:v]settb=AVTB,setpts=PTS-STARTPTS[main];[1:v]settb=AVTB,setpts=PTS-STARTPTS[ref];[main][ref]libvmaf=log_fmt=json:log_path=output.json" -f null -
This is the CUDA variant of the libvmaf filter. It only accepts CUDA frames.
It requires Netflix's vmaf library (libvmaf) as a pre-requisite. After installing the library it can be enabled using: "./configure --enable-nonfree --enable-ffnvcodec --enable-libvmaf".
Examples
ffmpeg \
-hwaccel cuda -hwaccel_output_format cuda -codec:v av1_cuvid -i dis.obu \
-hwaccel cuda -hwaccel_output_format cuda -codec:v av1_cuvid -i ref.obu \
-filter_complex "
[0:v]scale_cuda=format=yuv420p[ref]; \
[1:v]scale_cuda=format=yuv420p[dis]; \
[dis][ref]libvmaf_cuda=log_fmt=json:log_path=output.json
" \
-f null -
Apply limited difference filter using second and optionally third video stream.
The filter accepts the following options:
Commands
This filter supports the all above options as commands except option reference.
Limits the pixel components values to the specified range [min, max].
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Loop video frames.
The filter accepts the following options:
Examples
loop=loop=-1:size=1:start=0
loop=loop=10:size=1:start=0
loop=loop=5:size=10:start=0
Apply a 1D LUT to an input video.
The filter accepts the following options:
Currently supported formats:
Available values are:
Commands
This filter supports the all above options as commands.
Apply a 3D LUT to an input video.
The filter accepts the following options:
Currently supported formats:
Available values are:
Commands
This filter supports the "interp" option as commands.
Turn certain luma values into transparency.
The filter accepts the following options:
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Compute a look-up table for binding each pixel component input value to an output value, and apply it to the input video.
lutyuv applies a lookup table to a YUV input video, lutrgb to an RGB input video.
These filters accept the following parameters:
Each of them specifies the expression to use for computing the lookup table for the corresponding pixel component values.
The exact component associated to each of the c* options depends on the format in input.
The lut filter requires either YUV or RGB pixel formats in input, lutrgb requires RGB pixel formats in input, and lutyuv requires YUV.
The expressions can contain the following constants and functions:
All expressions default to "clipval".
Commands
This filter supports same commands as options.
Examples
lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
The above is the same as:
lutrgb="r=negval:g=negval:b=negval"
lutyuv="y=negval:u=negval:v=negval"
lutyuv=y=negval
lutyuv="u=128:v=128"
lutyuv="y=2*val"
lutrgb="g=0:b=0"
format=rgba,lutrgb=a="maxval-minval/2"
lutyuv=y=gammaval(0.5)
lutyuv=y='bitand(val, 128+64+32)'
lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
The "lut2" filter takes two input streams and outputs one stream.
The "tlut2" (time lut2) filter takes two consecutive frames from one single stream.
This filter accepts the following parameters:
The "lut2" filter also supports the framesync options.
Each of them specifies the expression to use for computing the lookup table for the corresponding pixel component values.
The exact component associated to each of the c* options depends on the format in inputs.
The expressions can contain the following constants:
All expressions default to "x".
Commands
This filter supports the all above options as commands except option "d".
Examples
lut2='ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,0,pow(2,bdx)-1)'
lut2='ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,pow(2,bdx-1),pow(2,bdx)-1):ifnot(x-y,pow(2,bdx-1),pow(2,bdx)-1)'
lut2='if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1))):if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1))):if(lt(x,y),0,if(gt(x,y),pow(2,bdx)-1,pow(2,bdx-1)))'
Clamp the first input stream with the second input and third input stream.
Returns the value of first stream to be between second input stream - "undershoot" and third input stream + "overshoot".
This filter accepts the following options:
Commands
This filter supports the all above options as commands.
Merge the second and third input stream into output stream using absolute differences between second input stream and first input stream and absolute difference between third input stream and first input stream. The picked value will be from second input stream if second absolute difference is greater than first one or from third input stream otherwise.
This filter accepts the following options:
Commands
This filter supports the all above options as commands.
Merge the first input stream with the second input stream using per pixel weights in the third input stream.
A value of 0 in the third stream pixel component means that pixel component from first stream is returned unchanged, while maximum value (eg. 255 for 8-bit videos) means that pixel component from second stream is returned unchanged. Intermediate values define the amount of merging between both input stream's pixel components.
This filter accepts the following options:
Commands
This filter supports the all above options as commands.
Merge the second and third input stream into output stream using absolute differences between second input stream and first input stream and absolute difference between third input stream and first input stream. The picked value will be from second input stream if second absolute difference is less than first one or from third input stream otherwise.
This filter accepts the following options:
Commands
This filter supports the all above options as commands.
Pick pixels comparing absolute difference of two video streams with fixed threshold.
If absolute difference between pixel component of first and second video stream is equal or lower than user supplied threshold than pixel component from first video stream is picked, otherwise pixel component from second video stream is picked.
This filter accepts the following options:
Commands
This filter supports the all above options as commands.
Create mask from input video.
For example it is useful to create motion masks after "tblend" filter.
This filter accepts the following options:
Commands
This filter supports the all above options as commands.
Apply motion-compensation deinterlacing.
It needs one field per frame as input and must thus be used together with yadif=1/3 or equivalent.
This filter accepts the following options:
It accepts one of the following values:
Default value is fast.
Default value is bff.
Higher values should result in a smoother motion vector field but less optimal individual vectors. Default value is 1.
Pick median pixel from certain rectangle defined by radius.
This filter accepts the following options:
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Merge color channel components from several video streams.
The filter accepts up to 4 input streams, and merge selected input planes to the output video.
This filter accepts the following options:
The mappings is specified as a bitmap. It should be specified as a hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the mapping for the first plane of the output stream. 'A' sets the number of the input stream to use (from 0 to 3), and 'a' the plane number of the corresponding input to use (from 0 to 3). The rest of the mappings is similar, 'Bb' describes the mapping for the output stream second plane, 'Cc' describes the mapping for the output stream third plane and 'Dd' describes the mapping for the output stream fourth plane.
Examples
[a0][a1][a2]mergeplanes=0x001020:yuv444p
[a0][a1]mergeplanes=0x00010210:yuva444p
format=yuva444p,mergeplanes=0x03010200:yuva444p
format=yuv420p,mergeplanes=0x000201:yuv420p
format=rgb24,mergeplanes=0x000102:yuv444p
Estimate and export motion vectors using block matching algorithms. Motion vectors are stored in frame side data to be used by other filters.
This filter accepts the following options:
Default value is esa.
Apply Midway Image Equalization effect using two video streams.
Midway Image Equalization adjusts a pair of images to have the same histogram, while maintaining their dynamics as much as possible. It's useful for e.g. matching exposures from a pair of stereo cameras.
This filter has two inputs and one output, which must be of same pixel format, but may be of different sizes. The output of filter is first input adjusted with midway histogram of both inputs.
This filter accepts the following option:
Convert the video to specified frame rate using motion interpolation.
This filter accepts the following options:
Default mode is obmc.
Default mode is bilat.
Default algorithm is epzs.
Default method is fdiff.
Mix several video input streams into one video stream.
A description of the accepted options follows.
Commands
This filter supports the following commands:
Convert video to gray using custom color filter.
A description of the accepted options follows.
Commands
This filter supports the all above options as commands.
This filter allows to apply main morphological grayscale transforms, erode and dilate with arbitrary structures set in second input stream.
Unlike naive implementation and much slower performance in erosion and dilation filters, when speed is critical "morpho" filter should be used instead.
A description of accepted options follows,
Default is "erode".
The "morpho" filter also supports the framesync options.
Commands
This filter supports same commands as options.
Drop frames that do not differ greatly from the previous frame in order to reduce frame rate.
The main use of this filter is for very-low-bitrate encoding (e.g. streaming over dialup modem), but it could in theory be used for fixing movies that were inverse-telecined incorrectly.
A description of the accepted options follows.
Default value is 0.
Default value is 0.
Values for hi and lo are for 8x8 pixel blocks and represent actual pixel value differences, so a threshold of 64 corresponds to 1 unit of difference for each pixel, or the same spread out differently over the block.
A frame is a candidate for dropping if no 8x8 blocks differ by more than a threshold of hi, and if no more than frac blocks (1 meaning the whole image) differ by more than a threshold of lo.
Default value for hi is 64*12, default value for lo is 64*5, and default value for frac is 0.33.
Obtain the MSAD (Mean Sum of Absolute Differences) between two input videos.
This filter takes two input videos.
Both input videos must have the same resolution and pixel format for this filter to work correctly. Also it assumes that both inputs have the same number of frames, which are compared one by one.
The obtained per component, average, min and max MSAD is printed through the logging system.
The filter stores the calculated MSAD of each frame in frame metadata.
This filter also supports the framesync options.
In the below example the input file main.mpg being processed is compared with the reference file ref.mpg.
ffmpeg -i main.mpg -i ref.mpg -lavfi msad -f null -
Multiply first video stream pixels values with second video stream pixels values.
The filter accepts the following options:
Commands
This filter supports same commands as options.
Negate (invert) the input video.
It accepts the following option:
Available values for components are:
Commands
This filter supports same commands as options.
Denoise frames using Non-Local Means algorithm.
Each pixel is adjusted by looking for other pixels with similar contexts. This context similarity is defined by comparing their surrounding patches of size pxp. Patches are searched in an area of rxr around the pixel.
Note that the research area defines centers for patches, which means some patches will be made of pixels outside that research area.
The filter accepts the following options.
The default value is 0 and means automatic.
The default value is 0 and means automatic.
Deinterlace video using neural network edge directed interpolation.
This filter accepts the following options:
Can be one of the following:
Can be one of the following:
Can be one of the following:
Default is "new".
Commands
This filter supports same commands as options, excluding weights option.
Force libavfilter not to use any of the specified pixel formats for the input to the next filter.
It accepts the following parameters:
Examples
noformat=pix_fmts=yuv420p,vflip
noformat=yuv420p|yuv444p|yuv410p
Add noise on video input frame.
The filter accepts the following options:
Examples
Add temporal and uniform noise to input video:
noise=alls=20:allf=t+u
Normalize RGB video (aka histogram stretching, contrast stretching). See: https://en.wikipedia.org/wiki/Normalization_(image_processing)
For each channel of each frame, the filter computes the input range and maps it linearly to the user-specified output range. The output range defaults to the full dynamic range from pure black to pure white.
Temporal smoothing can be used on the input range to reduce flickering (rapid changes in brightness) caused when small dark or bright objects enter or leave the scene. This is similar to the auto-exposure (automatic gain control) on a video camera, and, like a video camera, it may cause a period of over- or under-exposure of the video.
The R,G,B channels can be normalized independently, which may cause some color shifting, or linked together as a single channel, which prevents color shifting. Linked normalization preserves hue. Independent normalization does not, so it can be used to remove some color casts. Independent and linked normalization can be combined in any ratio.
The normalize filter accepts the following options:
Commands
This filter supports same commands as options, excluding smoothing option. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Examples
Stretch video contrast to use the full dynamic range, with no temporal smoothing; may flicker depending on the source content:
normalize=blackpt=black:whitept=white:smoothing=0
As above, but with 50 frames of temporal smoothing; flicker should be reduced, depending on the source content:
normalize=blackpt=black:whitept=white:smoothing=50
As above, but with hue-preserving linked channel normalization:
normalize=blackpt=black:whitept=white:smoothing=50:independence=0
As above, but with half strength:
normalize=blackpt=black:whitept=white:smoothing=50:independence=0:strength=0.5
Map the darkest input color to red, the brightest input color to cyan:
normalize=blackpt=red:whitept=cyan
Pass the video source unchanged to the output.
Optical Character Recognition
This filter uses Tesseract for optical character recognition. To enable compilation of this filter, you need to configure FFmpeg with "--enable-libtesseract".
It accepts the following options:
The filter exports recognized text as the frame metadata "lavfi.ocr.text". The filter exports confidence of recognized words as the frame metadata "lavfi.ocr.confidence".
Apply a video transform using libopencv.
To enable this filter, install the libopencv library and headers and configure FFmpeg with "--enable-libopencv".
It accepts the following parameters:
Refer to the official libopencv documentation for more precise information: <http://docs.opencv.org/master/modules/imgproc/doc/filtering.html>
Several libopencv filters are supported; see the following subsections.
dilate
Dilate an image by using a specific structuring element. It corresponds to the libopencv function "cvDilate".
It accepts the parameters: struct_el|nb_iterations.
struct_el represents a structuring element, and has the syntax: colsxrows+anchor_xxanchor_y/shape
cols and rows represent the number of columns and rows of the structuring element, anchor_x and anchor_y the anchor point, and shape the shape for the structuring element. shape must be "rect", "cross", "ellipse", or "custom".
If the value for shape is "custom", it must be followed by a string of the form "=filename". The file with name filename is assumed to represent a binary image, with each printable character corresponding to a bright pixel. When a custom shape is used, cols and rows are ignored, the number or columns and rows of the read file are assumed instead.
The default value for struct_el is "3x3+0x0/rect".
nb_iterations specifies the number of times the transform is applied to the image, and defaults to 1.
Some examples:
# Use the default values
ocv=dilate
# Dilate using a structuring element with a 5x5 cross, iterating two times
ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
# Read the shape from the file diamond.shape, iterating two times.
# The file diamond.shape may contain a pattern of characters like this
# *
# ***
# *****
# ***
# *
# The specified columns and rows are ignored
# but the anchor point coordinates are not
ocv=dilate:0x0+2x2/custom=diamond.shape|2
erode
Erode an image by using a specific structuring element. It corresponds to the libopencv function "cvErode".
It accepts the parameters: struct_el:nb_iterations, with the same syntax and semantics as the dilate filter.
smooth
Smooth the input video.
The filter takes the following parameters: type|param1|param2|param3|param4.
type is the type of smooth filter to apply, and must be one of the following values: "blur", "blur_no_scale", "median", "gaussian", or "bilateral". The default value is "gaussian".
The meaning of param1, param2, param3, and param4 depends on the smooth type. param1 and param2 accept integer positive values or 0. param3 and param4 accept floating point values.
The default value for param1 is 3. The default value for the other parameters is 0.
These parameters correspond to the parameters assigned to the libopencv function "cvSmooth".
2D Video Oscilloscope.
Useful to measure spatial impulse, step responses, chroma delays, etc.
It accepts the following parameters:
Commands
This filter supports same commands as options. The command accepts the same syntax of the corresponding option.
If the specified expression is not valid, it is kept at its current value.
Examples
oscilloscope=x=0.5:y=0:s=1
oscilloscope=x=0.5:y=1:s=1
oscilloscope=x=0.5:y=5/1080:s=1
oscilloscope=x=1:y=0.5:s=1:t=1
Overlay one video on top of another.
It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
It accepts the following parameters:
A description of the accepted options follows.
It accepts the following values:
Default value is frame.
It accepts the following values:
Default value is yuv420.
The x, and y expressions can contain the following parameters.
This filter also supports the framesync options.
Note that the n, t variables are available only when evaluation is done per frame, and will evaluate to NAN when eval is set to init.
Be aware that frames are taken from each input video in timestamp order, hence, if their initial timestamps differ, it is a good idea to pass the two inputs through a setpts=PTS-STARTPTS filter to have them begin in the same zero timestamp, as the example for the movie filter does.
You can chain together more overlays but you should test the efficiency of such approach.
Commands
This filter supports the following commands:
If the specified expression is not valid, it is kept at its current value.
Examples
overlay=main_w-overlay_w-10:main_h-overlay_h-10
Using named options the example above becomes:
overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
ffmpeg -i input -i logo1 -i logo2 -filter_complex 'overlay=x=10:y=H-h-10,overlay=x=W-w-10:y=H-h-10' output
color=color=red@.3:size=WxH [over]; [in][over] overlay [out]
ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
The above command is the same as:
ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
ffmpeg -i left.avi -i right.avi -filter_complex "
nullsrc=size=200x100 [background];
[0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
[1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
[background][left] overlay=shortest=1 [background+left];
[background+left][right] overlay=shortest=1:x=100 [left+right]
"
ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
-vf '[in]split[split_main][split_delogo];[split_delogo]trim=start=360:end=371,delogo=0:0:640:480[delogoed];[split_main][delogoed]overlay=eof_action=pass[out]'
masked.avi
nullsrc=s=200x200 [bg];
testsrc=s=100x100, split=4 [in0][in1][in2][in3];
[in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
[in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
[in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
[in3] null, [mid2] overlay=100:100 [out0]
Overlay one video on top of another.
This is the CUDA variant of the overlay filter. It only accepts CUDA frames. The underlying input pixel formats have to match.
It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
It accepts the following parameters:
They can contain the following parameters:
Default value is "0" for both expressions.
It accepts the following values:
Default value is frame.
This filter also supports the framesync options.
Apply Overcomplete Wavelet denoiser.
The filter accepts the following options:
Larger depth values will denoise lower frequency components more, but slow down filtering.
Must be an int in the range 8-16, default is 8.
Must be a double value in the range 0-1000, default is 1.0.
Must be a double value in the range 0-1000, default is 1.0.
Add paddings to the input image, and place the original input at the provided x, y coordinates.
It accepts the following parameters:
The width expression can reference the value set by the height expression, and vice versa.
The default value of width and height is 0.
The x expression can reference the value set by the y expression, and vice versa.
The default value of x and y is 0.
If x or y evaluate to a negative number, they'll be changed so the input image is centered on the padded area.
The default value of color is "black".
It accepts the following values:
Default value is init.
The value for the width, height, x, and y options are expressions containing the following constants:
Examples
pad=640:480:0:40:violet
The example above is equivalent to the following command:
pad=width=640:height=480:x=0:y=40:color=violet
pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
(ih * X / ih) * sar = output_dar
X = output_dar / sar
Thus the previous example needs to be modified to:
pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
pad="2*iw:2*ih:ow-iw:oh-ih"
Generate one palette for a whole video stream.
It accepts the following options:
It accepts the following values:
Default value is full.
The filter also exports the frame metadata "lavfi.color_quant_ratio" ("nb_color_in / nb_color_out") which you can use to evaluate the degree of color quantization of the palette. This information is also visible at info logging level.
Examples
ffmpeg -i input.mkv -vf palettegen palette.png
Use a palette to downsample an input video stream.
The filter takes two inputs: one video stream and a palette. The palette must be a 256 pixels image.
It accepts the following options:
Default is sierra2_4a.
The option must be an integer value in the range [0,5]. Default is 2.
Default is none.
The option must be an integer value in the range [0,255]. Default is 128.
Examples
ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
Correct perspective of video not recorded perpendicular to the screen.
A description of the accepted parameters follows.
The expressions can use the following variables:
It accepts the following values:
Default value is linear.
It accepts the following values:
Default value is source.
It accepts the following values:
Default value is init.
Delay interlaced video by one field time so that the field order changes.
The intended use is to fix PAL movies that have been captured with the opposite field order to the film-to-video transfer.
A description of the accepted parameters follows.
It accepts the following values:
Commands
This filter supports the all above options as commands.
Reduce various flashes in video, so to help users with epilepsy.
It accepts the following options:
Pixel format descriptor test filter, mainly useful for internal testing. The output video should be equal to the input video.
For example:
format=monow, pixdesctest
can be used to test the monowhite pixel format descriptor definition.
Apply pixelization to video stream.
The filter accepts the following options:
Possible values are:
Default value is "avg".
Commands
This filter supports all options as commands.
Display sample values of color channels. Mainly useful for checking color and levels. Minimum supported resolution is 640x480.
The filters accept the following options:
Commands
This filter supports same commands as options.
Enable the specified chain of postprocessing subfilters using libpostproc. This library should be automatically selected with a GPL build ("--enable-gpl"). Subfilters must be separated by '/' and can be disabled by prepending a '-'. Each subfilter and some options have a short and a long name that can be used interchangeably, i.e. dr/dering are the same.
The filters accept the following options:
All subfilters share common options to determine their scope:
These options can be appended after the subfilter name, separated by a '|'.
Available subfilters are:
The horizontal and vertical deblocking filters share the difference and flatness values so you cannot set different horizontal and vertical thresholds.
Examples
pp=hb/vb/dr/al
pp=de/-al
pp=default/tmpnoise|1|2|3
pp=hb|y/vb|a
Apply Postprocessing filter 7. It is variant of the spp filter, similar to spp = 6 with 7 point DCT, where only the center sample is used after IDCT.
The filter accepts the following options:
Apply alpha premultiply effect to input video stream using first plane of second stream as alpha.
Both streams must have same dimensions and same pixel format.
The filter accepts the following option:
Apply prewitt operator to input video stream.
The filter accepts the following option:
Commands
This filter supports the all above options as commands.
Alter frame colors in video with pseudocolors.
This filter accepts the following options:
Available LUTs:
Each of the expression options specifies the expression to use for computing the lookup table for the corresponding pixel component values.
The expressions can contain the following constants and functions:
All expressions default to "val".
Commands
This filter supports the all above options as commands.
Examples
pseudocolor="'if(between(val,ymax,amax),lerp(ymin,ymax,(val-ymax)/(amax-ymax)),-1):if(between(val,ymax,amax),lerp(umax,umin,(val-ymax)/(amax-ymax)),-1):if(between(val,ymax,amax),lerp(vmin,vmax,(val-ymax)/(amax-ymax)),-1):-1'"
Obtain the average, maximum and minimum PSNR (Peak Signal to Noise Ratio) between two input videos.
This filter takes in input two input videos, the first input is considered the "main" source and is passed unchanged to the output. The second input is used as a "reference" video for computing the PSNR.
Both video inputs must have the same resolution and pixel format for this filter to work correctly. Also it assumes that both inputs have the same number of frames, which are compared one by one.
The obtained average PSNR is printed through the logging system.
The filter stores the accumulated MSE (mean squared error) of each frame, and at the end of the processing it is averaged across all frames equally, and the following formula is applied to obtain the PSNR:
PSNR = 10*log10(MAX^2/MSE)
Where MAX is the average of the maximum values of each component of the image.
The description of the accepted parameters follows.
This filter also supports the framesync options.
The file printed if stats_file is selected, contains a sequence of key/value pairs of the form key:value for each compared couple of frames.
If a stats_version greater than 1 is specified, a header line precedes the list of per-frame-pair stats, with key value pairs following the frame format with the following parameters:
A description of each shown per-frame-pair parameter follows:
Examples
movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
[main][ref] psnr="stats_file=stats.log" [out]
On this example the input file being processed is compared with the reference file ref_movie.mpg. The PSNR of each individual frame is stored in stats.log.
ffmpeg -i main.mpg -i ref.mkv -lavfi "[0:v]settb=AVTB,setpts=PTS-STARTPTS[main];[1:v]settb=AVTB,setpts=PTS-STARTPTS[ref];[main][ref]psnr" -f null -
Pulldown reversal (inverse telecine) filter, capable of handling mixed hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive content.
The pullup filter is designed to take advantage of future context in making its decisions. This filter is stateless in the sense that it does not lock onto a pattern to follow, but it instead looks forward to the following fields in order to identify matches and rebuild progressive frames.
To produce content with an even framerate, insert the fps filter after pullup, use "fps=24000/1001" if the input frame rate is 29.97fps, "fps=24" for 30fps and the (rare) telecined 25fps input.
The filter accepts the following options:
This option may be set to use chroma plane instead of the default luma plane for doing filter's computations. This may improve accuracy on very clean source material, but more likely will decrease accuracy, especially if there is chroma noise (rainbow effect) or any grayscale video. The main purpose of setting mp to a chroma plane is to reduce CPU load and make pullup usable in realtime on slow machines.
For best results (without duplicated frames in the output file) it is necessary to change the output frame rate. For example, to inverse telecine NTSC input:
ffmpeg -i input -vf pullup -r 24000/1001 ...
Change video quantization parameters (QP).
The filter accepts the following option:
The expression is evaluated through the eval API and can contain, among others, the following constants:
Examples
qp=2+2*sin(PI*qp)
Flush video frames from internal cache of frames into a random order. No frame is discarded. Inspired by frei0r nervous filter.
Read closed captioning (EIA-608) information from the top lines of a video frame.
This filter adds frame metadata for "lavfi.readeia608.X.cc" and "lavfi.readeia608.X.line", where "X" is the number of the identified line with EIA-608 data (starting from 0). A description of each metadata value follows:
This filter accepts the following options:
Commands
This filter supports the all above options as commands.
Examples
ffprobe -f lavfi -i movie=captioned_video.mov,readeia608 -show_entries frame=pts_time:frame_tags=lavfi.readeia608.0.cc,lavfi.readeia608.1.cc -of csv
Read vertical interval timecode (VITC) information from the top lines of a video frame.
The filter adds frame metadata key "lavfi.readvitc.tc_str" with the timecode value, if a valid timecode has been detected. Further metadata key "lavfi.readvitc.found" is set to 0/1 depending on whether timecode data has been found or not.
This filter accepts the following options:
Examples
ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--}:x=(w-tw)/2:y=400-ascent'
Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
Destination pixel at position (X, Y) will be picked from source (x, y) position where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero value for pixel will be used for destination pixel.
Xmap and Ymap input video streams must be of same dimensions. Output video stream will have Xmap/Ymap video stream dimensions. Xmap and Ymap input video streams are 16bit depth, single channel.
The removegrain filter is a spatial denoiser for progressive video.
Range of mode is from 0 to 24. Description of each mode follows:
Suppress a TV station logo, using an image file to determine which pixels comprise the logo. It works by filling in the pixels that comprise the logo with neighboring pixels.
The filter accepts the following options:
Pixels in the provided bitmap image with a value of zero are not considered part of the logo, non-zero pixels are considered part of the logo. If you use white (255) for the logo and black (0) for the rest, you will be safe. For making the filter bitmap, it is recommended to take a screen capture of a black frame with the logo visible, and then using a threshold filter followed by the erode filter once or twice.
If needed, little splotches can be fixed manually. Remember that if logo pixels are not covered, the filter quality will be much reduced. Marking too many pixels as part of the logo does not hurt as much, but it will increase the amount of blurring needed to cover over the image and will destroy more information than necessary, and extra pixels will slow things down on a large logo.
This filter uses the repeat_field flag from the Video ES headers and hard repeats fields based on its value.
Reverse a video clip.
Warning: This filter requires memory to buffer the entire clip, so trimming is suggested.
Examples
trim=end=5,reverse
Shift R/G/B/A pixels horizontally and/or vertically.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Apply roberts cross operator to input video stream.
The filter accepts the following option:
Commands
This filter supports the all above options as commands.
Rotate video by an arbitrary angle expressed in radians.
The filter accepts the following options:
A description of the optional parameters follows.
This expression is evaluated for each frame.
Default value is "black".
The expressions for the angle and the output size can contain the following constants and functions:
These are only available when computing the out_w and out_h expressions.
Examples
rotate=PI/6
rotate=-PI/6
rotate=45*PI/180
rotate=PI/3+2*PI*t/T
rotate=A*sin(2*PI/T*t)
rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
Commands
The filter supports the following commands:
If the specified expression is not valid, it is kept at its current value.
Apply Shape Adaptive Blur.
The filter accepts the following options:
Each chroma option value, if not explicitly specified, is set to the corresponding luma option value.
Scale (resize) the input video, using the libswscale library.
The scale filter forces the output display aspect ratio to be the same of the input, by changing the output sample aspect ratio.
If the input image format is different from the format requested by the next filter, the scale filter will convert the input to the requested format.
Options
The filter accepts the following options, or any of the options supported by the libswscale scaler.
See the ffmpeg-scaler manual for the complete list of scaler options.
If the width or w value is 0, the input width is used for the output. If the height or h value is 0, the input height is used for the output.
If one and only one of the values is -n with n >= 1, the scale filter will use a value that maintains the aspect ratio of the input image, calculated from the other specified dimension. After that it will, however, make sure that the calculated dimension is divisible by n and adjust the value if necessary.
If both values are -n with n >= 1, the behavior will be identical to both values being set to 0 as previously detailed.
See below for the list of accepted constants for use in the dimension expression.
Default value is init.
Default value is 0.
This allows the autodetected value to be overridden as well as allows forcing a specific value used for the output and encoder.
If not specified, the color space type depends on the pixel format.
Possible values:
This allows the autodetected value to be overridden as well as allows forcing a specific value used for the output and encoder. If not specified, the range depends on the pixel format. Possible values:
One useful instance of this option is that when you know a specific device's maximum allowed resolution, you can use this to limit the output video to that, while retaining the aspect ratio. For example, device A allows 1280x720 playback, and your video is 1920x800. Using this option (set it to decrease) and specifying 1280x720 to the command line makes the output 1280x533.
Please note that this is a different thing than specifying -1 for w or h, you still need to specify the output resolution for this option to work.
This option respects the value set for force_original_aspect_ratio, increasing or decreasing the resolution accordingly. The video's aspect ratio may be slightly modified.
This option can be handy if you need to have a video fit within or exceed a defined resolution using force_original_aspect_ratio but also have encoder restrictions on width or height divisibility.
The values of the w and h options are expressions containing the following constants:
Examples
scale=w=200:h=100
This is equivalent to:
scale=200:100
or:
scale=200x100
scale=qcif
which can also be written as:
scale=size=qcif
scale=w=2*iw:h=2*ih
scale=2*in_w:2*in_h
scale=2*iw:2*ih:interl=1
scale=w=iw/2:h=ih/2
scale=3/2*iw:ow
scale=iw:1/PHI*iw
scale=ih*PHI:ih
scale=w=3/2*oh:h=3/5*ih
scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
scale=w='min(500\, iw*3/2):h=-1'
scale='trunc(ih*dar):ih',setsar=1/1
scale='trunc(ih*dar/2)*2:trunc(ih/2)*2',setsar=1/1
Commands
This filter supports the following commands:
Scale (resize) and convert (pixel format) the input video, using accelerated CUDA kernels. Setting the output width and height works in the same way as for the scale filter.
The filter accepts the following options:
Allows for the same expressions as the scale filter.
The filter does not support converting between YUV and RGB pixel formats.
If set to 1, frames are passed through as-is if they match the desired output parameters. This is the default behaviour.
Affects the curves of the bicubic algorithm.
Examples
scale_cuda=-2:720:format=yuv420p
scale_cuda=4096:2160:interp_algo=nearest
scale_cuda=passthrough=0
Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel format conversion on CUDA video frames. Setting the output width and height works in the same way as for the scale filter.
The following additional options are accepted:
One useful instance of this option is that when you know a specific device's maximum allowed resolution, you can use this to limit the output video to that, while retaining the aspect ratio. For example, device A allows 1280x720 playback, and your video is 1920x800. Using this option (set it to decrease) and specifying 1280x720 to the command line makes the output 1280x533.
Please note that this is a different thing than specifying -1 for w or h, you still need to specify the output resolution for this option to work.
This option respects the value set for force_original_aspect_ratio, increasing or decreasing the resolution accordingly. The video's aspect ratio may be slightly modified.
This option can be handy if you need to have a video fit within or exceed a defined resolution using force_original_aspect_ratio but also have encoder restrictions on width or height divisibility.
The values of the w and h options are expressions containing the following constants:
Scale (resize) the input video, based on a reference video.
See the scale filter for available options, scale2ref supports the same but uses the reference video instead of the main input as basis. scale2ref also supports the following additional constants for the w and h options:
Examples
'scale2ref[b][a];[a][b]overlay'
[logo-in][video-in]scale2ref=w=oh*mdar:h=ih/10[logo-out][video-out]
Commands
This filter supports the following commands:
Use the NVIDIA Performance Primitives (libnpp) to scale (resize) the input video, based on a reference video.
See the scale_npp filter for available options, scale2ref_npp supports the same but uses the reference video instead of the main input as basis. scale2ref_npp also supports the following additional constants for the w and h options:
Examples
'scale2ref_npp[b][a];[a][b]overlay_cuda'
[logo-in][video-in]scale2ref_npp=w=oh*mdar:h=ih/10[logo-out][video-out]
Scale and convert the color parameters using VTPixelTransferSession.
The filter accepts the following options:
Apply scharr operator to input video stream.
The filter accepts the following option:
Commands
This filter supports the all above options as commands.
Scroll input video horizontally and/or vertically by constant speed.
The filter accepts the following options:
Commands
This filter supports the following commands:
Detect video scene change.
This filter sets frame metadata with mafd between frame, the scene score, and forward the frame to the next filter, so they can use these metadata to detect scene change or others.
In addition, this filter logs a message and sets frame metadata when it detects a scene change by threshold.
"lavfi.scd.mafd" metadata keys are set with mafd for every frame.
"lavfi.scd.score" metadata keys are set with scene change score for every frame to detect scene change.
"lavfi.scd.time" metadata keys are set with current filtered frame time which detect scene change with threshold.
The filter accepts the following options:
Default value is 10..
Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined by the "purity" of the color (that is, how saturated it already is).
This filter is similar to the Adobe Photoshop Selective Color tool.
The filter accepts the following options:
Available values are:
Default is "absolute".
All the adjustment settings (reds, yellows, ...) accept up to 4 space separated floating point adjustment values in the [-1,1] range, respectively to adjust the amount of cyan, magenta, yellow and black for the pixels of its range.
Examples
selectivecolor=greens=.5 0 -.33 0:blues=0 .27
selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
The "separatefields" takes a frame-based video input and splits each frame into its components fields, producing a new half height clip with twice the frame rate and twice the frame count.
This filter use field-dominance information in frame to decide which of each pair of fields to place first in the output. If it gets it wrong use setfield filter before "separatefields" filter.
The "setdar" filter sets the Display Aspect Ratio for the filter output video.
This is done by changing the specified Sample (aka Pixel) Aspect Ratio, according to the following equation:
<DAR> = <HORIZONTAL_RESOLUTION> / <VERTICAL_RESOLUTION> * <SAR>
Keep in mind that the "setdar" filter does not modify the pixel dimensions of the video frame. Also, the display aspect ratio set by this filter may be changed by later filters in the filterchain, e.g. in case of scaling or if another "setdar" or a "setsar" filter is applied.
The "setsar" filter sets the Sample (aka Pixel) Aspect Ratio for the filter output video.
Note that as a consequence of the application of this filter, the output display aspect ratio will change according to the equation above.
Keep in mind that the sample aspect ratio set by the "setsar" filter may be changed by later filters in the filterchain, e.g. if another "setsar" or a "setdar" filter is applied.
It accepts the following parameters:
The parameter can be a floating point number string, or an expression. If the parameter is not specified, the value "0" is assumed, meaning that the same input value is used.
The parameter sar is an expression containing the following constants:
Examples
setdar=dar=1.77777
setdar=dar=16/9
setsar=sar=10/11
setdar=ratio=16/9:max=1000
Force field for the output video frame.
The "setfield" filter marks the interlace type field for the output frames. It does not change the input frame, but only sets the corresponding property, which affects how the frame is treated by following filters (e.g. "fieldorder" or "yadif").
The filter accepts the following options:
Force frame parameter for the output video frame.
The "setparams" filter marks interlace and color range for the output frames. It does not change the input frame, but only sets the corresponding property, which affects how the frame is treated by filters/encoders.
Use the NVIDIA Performance Primitives (libnpp) to perform image sharpening with border control.
The following additional options are accepted:
Apply shear transform to input video.
This filter supports the following options:
Default value is "black".
Commands
This filter supports the all above options as commands.
Show a line containing various information for each input video frame. The input video is not modified.
This filter supports the following options:
The shown line contains a sequence of key/value pairs of the form key:value.
The following values are shown in the output:
Displays the 256 colors palette of each frame. This filter is only relevant for pal8 pixel format frames.
It accepts the following option:
Reorder and/or duplicate and/or drop video frames.
It accepts the following parameters:
The first frame has the index 0. The default is to keep the input unchanged.
Examples
ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
Reorder pixels in video frames.
This filter accepts the following options:
Reorder and/or duplicate video planes.
It accepts the following parameters:
The first plane has the index 0. The default is to keep the input unchanged.
Examples
ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
Evaluate various visual metrics that assist in determining issues associated with the digitization of analog video media.
By default the filter will log these metadata values:
The filter accepts the following options:
Both options accept the following values:
Examples
ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
ffplay example.mov -vf signalstats="out=brng:color=red"
ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
The contents of signalstat_drawtext.txt used in the command are:
time %{pts:hms}
Y (%{metadata:lavfi.signalstats.YMIN}-%{metadata:lavfi.signalstats.YMAX})
U (%{metadata:lavfi.signalstats.UMIN}-%{metadata:lavfi.signalstats.UMAX})
V (%{metadata:lavfi.signalstats.VMIN}-%{metadata:lavfi.signalstats.VMAX})
saturation maximum: %{metadata:lavfi.signalstats.SATMAX}
Calculates the MPEG-7 Video Signature. The filter can handle more than one input. In this case the matching between the inputs can be calculated additionally. The filter always passes through the first input. The signature of each stream can be written into a file.
It accepts the following options:
Available values are:
Available values are:
Examples
ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
ffmpeg -i input1.mkv -i input2.mkv -filter_complex "[0:v][1:v] signature=nb_inputs=2:detectmode=full:format=xml:filename=signature%d.xml" -map :v -f null -
Calculate Spatial Information (SI) and Temporal Information (TI) scores for a video, as defined in ITU-T Rec. P.910 (11/21): Subjective video quality assessment methods for multimedia applications. Available PDF at <https://www.itu.int/rec/T-REC-P.910-202111-S/en>. Note that this is a legacy implementation that corresponds to a superseded recommendation. Refer to ITU-T Rec. P.910 (07/22) for the latest version: <https://www.itu.int/rec/T-REC-P.910-202207-I/en>
It accepts the following option:
Examples
ffmpeg -i input.mp4 -vf siti=print_summary=1 -f null -
Blur the input video without impacting the outlines.
It accepts the following options:
If a chroma option is not explicitly set, the corresponding luma value is set.
Apply sobel operator to input video stream.
The filter accepts the following option:
Commands
This filter supports the all above options as commands.
Apply a simple postprocessing filter that compresses and decompresses the image at several (or - in the case of quality level 6 - all) shifts and average the results.
The filter accepts the following options:
Commands
This filter supports the following commands:
Scale the input by applying one of the super-resolution methods based on convolutional neural networks. Supported models:
Training scripts as well as scripts for model file (.pb) saving can be found at <https://github.com/XueweiMeng/sr/tree/sr_dnn_native>. Original repository is at <https://github.com/HighVoltageRocknRoll/sr.git>.
The filter accepts the following options:
To get full functionality (such as async execution), please use the dnn_processing filter.
Obtain the SSIM (Structural SImilarity Metric) between two input videos.
This filter takes in input two input videos, the first input is considered the "main" source and is passed unchanged to the output. The second input is used as a "reference" video for computing the SSIM.
Both video inputs must have the same resolution and pixel format for this filter to work correctly. Also it assumes that both inputs have the same number of frames, which are compared one by one.
The filter stores the calculated SSIM of each frame.
The description of the accepted parameters follows.
The file printed if stats_file is selected, contains a sequence of key/value pairs of the form key:value for each compared couple of frames.
A description of each shown parameter follows:
This filter also supports the framesync options.
Examples
movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
[main][ref] ssim="stats_file=stats.log" [out]
On this example the input file being processed is compared with the reference file ref_movie.mpg. The SSIM of each individual frame is stored in stats.log.
ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
ffmpeg -i main.mpg -i ref.mkv -lavfi "[0:v]settb=AVTB,setpts=PTS-STARTPTS[main];[1:v]settb=AVTB,setpts=PTS-STARTPTS[ref];[main][ref]ssim" -f null -
Convert between different stereoscopic image formats.
The filters accept the following options:
Available values for input image formats are:
Default value is sbsl.
Default value is arcd.
Examples
stereo3d=sbsl:aybd
stereo3d=abl:sbsr
Select video or audio streams.
The filter accepts the following options:
Commands
The "streamselect" and "astreamselect" filter supports the following commands:
Examples
sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
Draw subtitles on top of input video using the libass library.
To enable compilation of this filter you need to configure FFmpeg with "--enable-libass". This filter also requires a build with libavcodec and libavformat to convert the passed subtitles file to ASS (Advanced Substation Alpha) subtitles format.
The filter accepts the following options:
The option is enabled by default except for native ASS.
If the first key is not specified, it is assumed that the first value specifies the filename.
For example, to render the file sub.srt on top of the input video, use the command:
subtitles=sub.srt
which is equivalent to:
subtitles=filename=sub.srt
To render the default subtitles stream from file video.mkv, use:
subtitles=video.mkv
To render the second subtitles stream from that file, use:
subtitles=video.mkv:si=1
To make the subtitles stream from sub.srt appear in 80% transparent blue "DejaVu Serif", use:
subtitles=sub.srt:force_style='Fontname=DejaVu Serif,PrimaryColour=&HCCFF0000'
Scale the input by 2x and smooth using the Super2xSaI (Scale and Interpolate) pixel art scaling algorithm.
Useful for enlarging pixel art images without reducing sharpness.
Swap two rectangular objects in video.
This filter accepts the following options:
All expressions are evaluated once for each frame.
The all options are expressions containing the following constants:
Commands
This filter supports the all above options as commands.
Swap U & V plane.
Blend successive video frames.
See blend
Apply telecine process to the video.
This filter accepts the following options:
Some typical patterns:
NTSC output (30i):
27.5p: 32222
24p: 23 (classic)
24p: 2332 (preferred)
20p: 33
18p: 334
16p: 3444
PAL output (25i):
27.5p: 12222
24p: 222222222223 ("Euro pulldown")
16.67p: 33
16p: 33333334
Compute and draw a color distribution histogram for the input video across time.
Unlike histogram video filter which only shows histogram of single input frame at certain time, this filter shows also past histograms of number of frames defined by "width" option.
The computed histogram is a representation of the color component distribution in an image.
The filter accepts the following options:
Default is "stack".
Available values for slide is:
Default is "replace".
Apply threshold effect to video stream.
This filter needs four video streams to perform thresholding. First stream is stream we are filtering. Second stream is holding threshold values, third stream is holding min values, and last, fourth stream is holding max values.
The filter accepts the following option:
For example if first stream pixel's component value is less then threshold value of pixel component from 2nd threshold stream, third stream value will picked, otherwise fourth stream pixel component value will be picked.
Using color source filter one can perform various types of thresholding:
Commands
This filter supports the all options as commands.
Examples
ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
Select the most representative frame in a given sequence of consecutive frames.
The filter accepts the following options:
Since the filter keeps track of the whole frames sequence, a bigger n value will result in a higher memory usage, so a high value is not recommended.
Examples
thumbnail=50
ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
Tile several successive frames together.
The untile filter can do the reverse.
The filter accepts the following options:
Examples
ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
The -vsync 0 is necessary to prevent ffmpeg from duplicating each output frame to accommodate the originally detected frame rate.
tile=3x2:nb_frames=5:padding=7:margin=2
Perform various types of temporal field interlacing.
Frames are counted starting from 1, so the first input frame is considered odd.
The filter accepts the following options:
Available values are:
------> time
Input:
Frame 1 Frame 2 Frame 3 Frame 4
11111 22222 33333 44444
11111 22222 33333 44444
11111 22222 33333 44444
11111 22222 33333 44444
Output:
11111 33333
22222 44444
11111 33333
22222 44444
11111 33333
22222 44444
11111 33333
22222 44444
------> time
Input:
Frame 1 Frame 2 Frame 3 Frame 4
11111 22222 33333 44444
11111 22222 33333 44444
11111 22222 33333 44444
11111 22222 33333 44444
Output:
11111 33333
11111 33333
11111 33333
11111 33333
------> time
Input:
Frame 1 Frame 2 Frame 3 Frame 4
11111 22222 33333 44444
11111 22222 33333 44444
11111 22222 33333 44444
11111 22222 33333 44444
Output:
22222 44444
22222 44444
22222 44444
22222 44444
------> time
Input:
Frame 1 Frame 2 Frame 3 Frame 4
11111 22222 33333 44444
11111 22222 33333 44444
11111 22222 33333 44444
11111 22222 33333 44444
Output:
11111 ..... 33333 .....
..... 22222 ..... 44444
11111 ..... 33333 .....
..... 22222 ..... 44444
11111 ..... 33333 .....
..... 22222 ..... 44444
11111 ..... 33333 .....
..... 22222 ..... 44444
------> time
Input:
Frame 1 Frame 2 Frame 3 Frame 4
11111<- 22222 33333<- 44444
11111 22222<- 33333 44444<-
11111<- 22222 33333<- 44444
11111 22222<- 33333 44444<-
Output:
11111 33333
22222 44444
11111 33333
22222 44444
------> time
Input:
Frame 1 Frame 2 Frame 3 Frame 4
11111 22222<- 33333 44444<-
11111<- 22222 33333<- 44444
11111 22222<- 33333 44444<-
11111<- 22222 33333<- 44444
Output:
22222 44444
11111 33333
22222 44444
11111 33333
------> time
Input:
Frame 1 Frame 2 Frame 3 Frame 4
11111 22222 33333 44444
11111 22222 33333 44444
11111 22222 33333 44444
11111 22222 33333 44444
Output:
11111 22222 22222 33333 33333 44444 44444
11111 11111 22222 22222 33333 33333 44444
11111 22222 22222 33333 33333 44444 44444
11111 11111 22222 22222 33333 33333 44444
------> time
Input:
Frame 1 Frame 2 Frame 3 Frame 4
11111 22222 33333 44444
11111 22222 33333 44444
11111 22222 33333 44444
11111 22222 33333 44444
Output:
11111 33333 33333 55555
22222 22222 44444 44444
11111 33333 33333 55555
22222 22222 44444 44444
11111 33333 33333 55555
22222 22222 44444 44444
11111 33333 33333 55555
22222 22222 44444 44444
Numeric values are deprecated but are accepted for backward compatibility reasons.
Default mode is "merge".
Available value for flags is:
Vertical low-pass filtering and bypassing already interlaced frames can only be enabled for mode interleave_top and interleave_bottom.
Pick median pixels from several successive input video frames.
The filter accepts the following options:
Commands
This filter supports all above options as commands, excluding option "radius".
Apply Temporal Midway Video Equalization effect.
Midway Video Equalization adjusts a sequence of video frames to have the same histograms, while maintaining their dynamics as much as possible. It's useful for e.g. matching exposures from a video frames sequence.
This filter accepts the following option:
Mix successive video frames.
A description of the accepted options follows.
Examples
tmix=frames=7:weights="1 1 1 1 1 1 1"
tmix=frames=3:weights="-1 3 -1"
tmix=frames=3:weights="-1 2 -1":scale=1
Commands
This filter supports the following commands:
Tone map colors from different dynamic ranges.
This filter expects data in single precision floating point, as it needs to operate on (and can output) out-of-range values. Another filter, such as zscale, is needed to convert the resulting frame to a usable format.
The tonemapping algorithms implemented only work on linear light, so input data should be linearized beforehand (and possibly correctly tagged).
ffmpeg -i INPUT -vf zscale=transfer=linear,tonemap=clip,zscale=transfer=bt709,format=yuv420p OUTPUT
Options
The filter accepts the following options.
Possible values are:
Default is none.
This affects the following algorithms:
The default of 2.0 is somewhat conservative and will mostly just apply to skies or directly sunlit surfaces. A setting of 0.0 disables this option.
This option works only if the input frame has a supported color tag.
Temporarily pad video frames.
The filter accepts the following options:
The default value of color is "black".
Transpose rows with columns in the input video and optionally flip it.
It accepts the following parameters:
Can assume the following values:
L.R L.l
. . -> . .
l.r R.r
L.R l.L
. . -> . .
l.r r.R
L.R R.r
. . -> . .
l.r L.l
L.R r.R
. . -> . .
l.r l.L
For values between 4-7, the transposition is only done if the input video geometry is portrait and not landscape. These values are deprecated, the "passthrough" option should be used instead.
Numerical values are deprecated, and should be dropped in favor of symbolic constants.
Default value is "none".
For example to rotate by 90 degrees clockwise and preserve portrait layout:
transpose=dir=1:passthrough=portrait
The command above can also be specified as:
transpose=1:portrait
Transpose rows with columns in the input video and optionally flip it. For more in depth examples see the transpose video filter, which shares mostly the same options.
It accepts the following parameters:
Can assume the following values:
Trim the input so that the output contains one continuous subpart of the input.
It accepts the following parameters:
start, end, and duration are expressed as time duration specifications; see the Time duration section in the ffmpeg-utils(1) manual for the accepted syntax.
Note that the first two sets of the start/end options and the duration option look at the frame timestamp, while the _frame variants simply count the frames that pass through the filter. Also note that this filter does not modify the timestamps. If you wish for the output timestamps to start at zero, insert a setpts filter after the trim filter.
If multiple start or end options are set, this filter tries to be greedy and keep all the frames that match at least one of the specified constraints. To keep only the part that matches all the constraints at once, chain multiple trim filters.
The defaults are such that all the input is kept. So it is possible to set e.g. just the end values to keep everything before the specified time.
Examples:
ffmpeg -i INPUT -vf trim=60:120
ffmpeg -i INPUT -vf trim=duration=1
Apply alpha unpremultiply effect to input video stream using first plane of second stream as alpha.
Both streams must have same dimensions and same pixel format.
The filter accepts the following option:
If the format has 1 or 2 components, then luma is bit 0. If the format has 3 or 4 components: for RGB formats bit 0 is green, bit 1 is blue and bit 2 is red; for YUV formats bit 0 is luma, bit 1 is chroma-U and bit 2 is chroma-V. If present, the alpha channel is always the last bit.
Sharpen or blur the input video.
It accepts the following parameters:
Negative values will blur the input video, while positive values will sharpen it, a value of zero will disable the effect.
Default value is 1.0.
Negative values will blur the input video, while positive values will sharpen it, a value of zero will disable the effect.
Default value is 0.0.
Negative values will blur the input video, while positive values will sharpen it, a value of zero will disable the effect.
Default value is 0.0.
All parameters are optional and default to the equivalent of the string '5:5:1.0:5:5:0.0'.
Examples
unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
unsharp=7:7:-2:7:7:-2
Decompose a video made of tiled images into the individual images.
The frame rate of the output video is the frame rate of the input video multiplied by the number of tiles.
This filter does the reverse of tile.
The filter accepts the following options:
Examples
ffmpeg -r 1 -i image.jpg -vf untile=1x25 movie.mkv
Apply ultra slow/simple postprocessing filter that compresses and decompresses the image at several (or - in the case of quality level 8 - all) shifts and average the results.
The way this differs from the behavior of spp is that uspp actually encodes & decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8 DCT similar to MJPEG.
This filter is only available in ffmpeg version 4.4 or earlier.
The filter accepts the following options:
Convert 360 videos between various formats.
The filter accepts the following options:
Format specific options:
Example values:
Default value is @samp{0}. Maximum value is @samp{0.1}.
Default value is @samp{0}. If greater than zero it overrides other padding options.
Designation of directions:
Default value is @samp{rludfb}.
Designation of angles:
Default value is @samp{000000}.
Format specific options:
If diagonal field of view is set it overrides horizontal and vertical field of view.
If diagonal field of view is set it overrides horizontal and vertical field of view.
Format specific options:
If diagonal field of view is set it overrides horizontal and vertical field of view.
If diagonal field of view is set it overrides horizontal and vertical field of view.
Format specific options:
If diagonal field of view is set it overrides horizontal and vertical field of view.
If diagonal field of view is set it overrides horizontal and vertical field of view.
Format specific options:
If diagonal field of view is set it overrides horizontal and vertical field of view.
If diagonal field of view is set it overrides horizontal and vertical field of view.
Format specific options:
Format specific options:
If diagonal field of view is set it overrides horizontal and vertical field of view.
If diagonal field of view is set it overrides horizontal and vertical field of view.
Format specific options:
Format specific options:
If diagonal field of view is set it overrides horizontal and vertical field of view.
If diagonal field of view is set it overrides horizontal and vertical field of view.
Format specific options:
If diagonal field of view is set it overrides horizontal and vertical field of view.
If diagonal field of view is set it overrides horizontal and vertical field of view.
Available methods:
Default value is @samp{line}.
Default resolution depends on formats.
Default value is @samp{2d} for input and output format.
Default value is @samp{ypr}.
Examples
ffmpeg -i input.mkv -vf v360=e:c3x2:cubic:out_pad=0.01 output.mkv
ffmpeg -i input.mkv -vf v360=eac:flat:yaw=180 output.mkv
v360=eac:equirect:in_stereo=sbs:in_trans=1:ih_flip=1:out_stereo=tb
Commands
This filter supports subset of above options as commands.
Apply a wavelet based denoiser.
It transforms each frame from the video input into the wavelet domain, using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to the obtained coefficients. It does an inverse wavelet transform after. Due to wavelet properties, it should give a nice smoothed result, and reduced noise, without blurring picture features.
This filter accepts the following options:
It accepts the following values:
Default is garrote.
It accepts the following values:
Default is universal.
Apply variable blur filter by using 2nd video stream to set blur radius. The 2nd stream must have the same dimensions.
This filter accepts the following options:
The "varblur" filter also supports the framesync options.
Commands
This filter supports all the above options as commands.
Display 2 color component values in the two dimensional graph (which is called a vectorscope).
This filter accepts the following options:
It accepts the following values:
Default is auto.
Analyze video stabilization/deshaking. Perform pass 1 of 2, see vidstabtransform for pass 2.
This filter generates a file with relative translation and rotation transform information about subsequent frames, which is then used by the vidstabtransform filter.
To enable compilation of this filter you need to configure FFmpeg with "--enable-libvidstab".
This filter accepts the following options:
If enabled, the motion of the frames is compared to a reference frame in the filtered stream, identified by the specified number. The idea is to compensate all movements in a more-or-less static scene and keep the camera view absolutely still.
If set to 0, it is disabled. The frames are counted starting from 1.
Examples
vidstabdetect
vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
vidstabdetect=show=1
ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
Video stabilization/deshaking: pass 2 of 2, see vidstabdetect for pass 1.
Read a file with transform information for each frame and apply/compensate them. Together with the vidstabdetect filter this can be used to deshake videos. See also <http://public.hronopik.de/vid.stab>. It is important to also use the unsharp filter, see below.
To enable compilation of this filter you need to configure FFmpeg with "--enable-libvidstab".
Options
For example a number of 10 means that 21 frames are used (10 in the past and 10 in the future) to smoothen the motion in the video. A larger value leads to a smoother video, but limits the acceleration of the camera (pan/tilt movements). 0 is a special case where a static camera is simulated.
Accepted values are:
Available values are:
Accepted values are:
Note that the value given at zoom is added to the one calculated here.
Available values are:
Use also "tripod" option of vidstabdetect.
Examples
ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
Note the use of the unsharp filter which is always recommended.
vidstabtransform=zoom=5:input="mytransforms.trf"
vidstabtransform=smoothing=30
Flip the input video vertically.
For example, to vertically flip a video with ffmpeg:
ffmpeg -i in.avi -vf "vflip" out.avi
Detect variable frame rate video.
This filter tries to detect if the input is variable or constant frame rate.
At end it will output number of frames detected as having variable delta pts, and ones with constant delta pts. If there was frames with variable delta, than it will also show min, max and average delta encountered.
Boost or alter saturation.
The filter accepts the following options:
Commands
This filter supports the all above options as commands.
Obtain the average VIF (Visual Information Fidelity) between two input videos.
This filter takes two input videos.
Both input videos must have the same resolution and pixel format for this filter to work correctly. Also it assumes that both inputs have the same number of frames, which are compared one by one.
The obtained average VIF score is printed through the logging system.
The filter stores the calculated VIF score of each frame.
This filter also supports the framesync options.
In the below example the input file main.mpg being processed is compared with the reference file ref.mpg.
ffmpeg -i main.mpg -i ref.mpg -lavfi vif -f null -
Make or reverse a natural vignetting effect.
The filter accepts the following options:
The value is clipped in the "[0,PI/2]" range.
Default value: "PI/5"
Available modes are:
Default value is forward.
It accepts the following values:
Default value is init.
Default is "1/1".
Expressions
The alpha, x0 and y0 expressions can contain the following parameters.
Examples
vignette=PI/4
vignette='PI/4+random(1)*PI/50':eval=frame
Obtain the average VMAF motion score of a video. It is one of the component metrics of VMAF.
The obtained average motion score is printed through the logging system.
The filter accepts the following options:
Example:
ffmpeg -i ref.mpg -vf vmafmotion -f null -
Stack input videos vertically.
All streams must be of same pixel format and of same width.
Note that this filter is faster than using overlay and pad filter to create same output.
The filter accepts the following options:
Deinterlace the input video ("w3fdif" stands for "Weston 3 Field Deinterlacing Filter").
Based on the process described by Martin Weston for BBC R&D, and implemented based on the de-interlace algorithm written by Jim Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter uses filter coefficients calculated by BBC R&D.
This filter uses field-dominance information in frame to decide which of each pair of fields to place first in the output. If it gets it wrong use setfield filter before "w3fdif" filter.
There are two sets of filter coefficients, so called "simple" and "complex". Which set of filter coefficients is used can be set by passing an optional parameter:
Default value is complex.
The default value is "field".
The default value is "auto". If the interlacing is unknown or the decoder does not export this information, top field first will be assumed.
Default value is all.
Commands
This filter supports same commands as options.
Video waveform monitor.
The waveform monitor plots color component intensity. By default luma only. Each column of the waveform corresponds to a column of pixels in the source video.
It accepts the following options:
This display mode makes it easier to spot relative differences or similarities in overlapping areas of the color components that are supposed to be identical, such as neutral whites, grays, or blacks.
Using this display mode makes it easy to spot color casts in the highlights and shadows of an image, by comparing the contours of the top and the bottom graphs of each waveform. Since whites, grays, and blacks are characterized by exactly equal amounts of red, green, and blue, neutral areas of the picture should display three waveforms of roughly equal width/height. If not, the correction is easy to perform by making level adjustments the three waveforms.
Default is "stack".
Default is digital.
Default is none.
The "weave" takes a field-based video input and join each two sequential fields into single frame, producing a new double height clip with half the frame rate and half the frame count.
The "doubleweave" works same as "weave" but without halving frame rate and frame count.
It accepts the following option:
Examples
separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
Apply the xBR high-quality magnification filter which is designed for pixel art. It follows a set of edge-detection rules, see <https://forums.libretro.com/t/xbr-algorithm-tutorial/123>.
It accepts the following option:
Apply normalized cross-correlation between first and second input video stream.
Second input video stream dimensions must be lower than first input video stream.
The filter accepts the following options:
The "xcorrelate" filter also supports the framesync options.
Apply cross fade from one input video stream to another input video stream. The cross fade is applied for specified duration.
Both inputs must be constant frame-rate and have the same resolution, pixel format, frame rate and timebase.
The filter accepts the following options:
Default transition effect is fade.
The expressions can use the following variables and functions:
Examples
ffmpeg -i first.mp4 -i second.mp4 -filter_complex xfade=transition=fade:duration=2:offset=5 output.mp4
Pick median pixels from several input videos.
The filter accepts the following options:
Commands
This filter supports all above options as commands, excluding option "inputs".
Stack video inputs into custom layout.
All streams must be of same pixel format.
The filter accepts the following options:
Note that if inputs are of different sizes gaps may appear, as not all of the output video frame will be filled. Similarly, videos can overlap each other if their position doesn't leave enough space for the full frame of adjoining videos.
For 2 inputs, a default layout of "0_0|w0_0" (equivalent to "grid=2x1") is set. In all other cases, a layout or a grid must be set by the user. Either "grid" or "layout" can be specified at a time. Specifying both will result in an error.
If "grid" is set, then "inputs" option is ignored and is implicitly set to "ROWS * COLUMNS".
For 2 inputs, a default grid of "2x1" (equivalent to "layout=0_0|w0_0") is set. In all other cases, a layout or a grid must be set by the user. Either "grid" or "layout" can be specified at a time. Specifying both will result in an error.
Examples
Layout:
input1(0, 0) | input3(w0, 0)
input2(0, h0) | input4(w0, h0)
xstack=inputs=4:layout=0_0|0_h0|w0_0|w0_h0
Note that if inputs are of different sizes, gaps or overlaps may occur.
Layout:
input1(0, 0)
input2(0, h0)
input3(0, h0+h1)
input4(0, h0+h1+h2)
xstack=inputs=4:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2
Note that if inputs are of different widths, unused space will appear.
Layout:
input1(0, 0) | input4(w0, 0) | input7(w0+w3, 0)
input2(0, h0) | input5(w0, h0) | input8(w0+w3, h0)
input3(0, h0+h1) | input6(w0, h0+h1) | input9(w0+w3, h0+h1)
xstack=inputs=9:layout=0_0|0_h0|0_h0+h1|w0_0|w0_h0|w0_h0+h1|w0+w3_0|w0+w3_h0|w0+w3_h0+h1
Note that if inputs are of different sizes, gaps or overlaps may occur.
Layout:
input1(0, 0) | input5(w0, 0) | input9 (w0+w4, 0) | input13(w0+w4+w8, 0)
input2(0, h0) | input6(w0, h0) | input10(w0+w4, h0) | input14(w0+w4+w8, h0)
input3(0, h0+h1) | input7(w0, h0+h1) | input11(w0+w4, h0+h1) | input15(w0+w4+w8, h0+h1)
input4(0, h0+h1+h2)| input8(w0, h0+h1+h2)| input12(w0+w4, h0+h1+h2)| input16(w0+w4+w8, h0+h1+h2)
xstack=inputs=16:layout=0_0|0_h0|0_h0+h1|0_h0+h1+h2|w0_0|w0_h0|w0_h0+h1|w0_h0+h1+h2|w0+w4_0|
w0+w4_h0|w0+w4_h0+h1|w0+w4_h0+h1+h2|w0+w4+w8_0|w0+w4+w8_h0|w0+w4+w8_h0+h1|w0+w4+w8_h0+h1+h2
Note that if inputs are of different sizes, gaps or overlaps may occur.
Deinterlace the input video ("yadif" means "yet another deinterlacing filter").
It accepts the following parameters:
The default value is "send_frame".
The default value is "auto". If the interlacing is unknown or the decoder does not export this information, top field first will be assumed.
The default value is "all".
Deinterlace the input video using the yadif algorithm, but implemented in CUDA so that it can work as part of a GPU accelerated pipeline with nvdec and/or nvenc.
It accepts the following parameters:
The default value is "send_frame".
The default value is "auto". If the interlacing is unknown or the decoder does not export this information, top field first will be assumed.
The default value is "all".
Apply blur filter while preserving edges ("yaepblur" means "yet another edge preserving blur filter"). The algorithm is described in "J. S. Lee, Digital image enhancement and noise filtering by use of local statistics, IEEE Trans. Pattern Anal. Mach. Intell. PAMI-2, 1980."
It accepts the following parameters:
Commands
This filter supports same commands as options.
Apply Zoom & Pan effect.
This filter accepts the following options:
Each expression can contain the following constants:
Examples
zoompan=z='min(zoom+0.0015,1.5)':d=700:x='if(gte(zoom,1.5),x,x+1/a)':y='if(gte(zoom,1.5),y,y+1)':s=640x360
zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
zoompan=z='if(between(in_time,0,1),2,1)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
Scale (resize) the input video, using the z.lib library: <https://github.com/sekrit-twc/zimg>. To enable compilation of this filter, you need to configure FFmpeg with "--enable-libzimg".
The zscale filter forces the output display aspect ratio to be the same as the input, by changing the output sample aspect ratio.
If the input image format is different from the format requested by the next filter, the zscale filter will convert the input to the requested format.
Options
The filter accepts the following options.
If the width or w value is 0, the input width is used for the output. If the height or h value is 0, the input height is used for the output.
If one and only one of the values is -n with n >= 1, the zscale filter will use a value that maintains the aspect ratio of the input image, calculated from the other specified dimension. After that it will, however, make sure that the calculated dimension is divisible by n and adjust the value if necessary.
If both values are -n with n >= 1, the behavior will be identical to both values being set to 0 as previously detailed.
See below for the list of accepted constants for use in the dimension expression.
Possible values are:
Default is none.
Possible values are:
Default is bilinear.
Possible values are:
Default is same as input.
Possible values are:
Default is same as input.
Possible values are:
Default is same as input.
Possible value are:
Default is same as input.
Possible values are:
Default is same as input.
Possible values are:
Default is same as input.
Possible values are:
Default is same as input.
Possible value are:
Possible values are:
Possible values are:
The values of the w and h options are expressions containing the following constants:
Commands
This filter supports the following commands:
Below is a description of the currently available OpenCL video filters.
To enable compilation of these filters you need to configure FFmpeg with "--enable-opencl".
Running OpenCL filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
For more detailed information see <https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options>
-init_hw_device opencl=gpu:1.0 -filter_hw_device gpu -i INPUT -vf "hwupload, avgblur_opencl, hwdownload" OUTPUT
Since OpenCL filters are not able to access frame data in normal memory, all frame data needs to be uploaded(hwupload) to hardware surfaces connected to the appropriate device before being used and then downloaded(hwdownload) back to normal memory. Note that hwupload will upload to a surface with the same layout as the software frame, so it may be necessary to add a format filter immediately before to get the input into the right format and hwdownload does not support all formats on the output - it may be necessary to insert an additional format filter immediately following in the graph to get the output in a supported format.
Apply average blur filter.
The filter accepts the following options:
Example
-i INPUT -vf "hwupload, avgblur_opencl=3, hwdownload" OUTPUT
Apply a boxblur algorithm to the input video.
It accepts the following parameters:
A description of the accepted options follows.
The radius value must be a non-negative number, and must not be greater than the value of the expression "min(w,h)/2" for the luma and alpha planes, and of "min(cw,ch)/2" for the chroma planes.
Default value for luma_radius is "2". If not specified, chroma_radius and alpha_radius default to the corresponding value set for luma_radius.
The expressions can contain the following constants:
Default value for luma_power is 2. If not specified, chroma_power and alpha_power default to the corresponding value set for luma_power.
A value of 0 will disable the effect.
Examples
Apply boxblur filter, setting each pixel of the output to the average value of box-radiuses luma_radius, chroma_radius, alpha_radius for each plane respectively. The filter will apply luma_power, chroma_power, alpha_power times onto the corresponding plane. For pixels on the edges of the image, the radius does not extend beyond the image boundaries, and so out-of-range coordinates are not used in the calculations.
-i INPUT -vf "hwupload, boxblur_opencl=luma_radius=2:luma_power=3, hwdownload" OUTPUT
-i INPUT -vf "hwupload, boxblur_opencl=2:3, hwdownload" OUTPUT
For the luma plane, a 2x2 box radius will be run once.
For the chroma plane, a 4x4 box radius will be run 5 times.
For the alpha plane, a 3x3 box radius will be run 7 times.
-i INPUT -vf "hwupload, boxblur_opencl=2:1:4:5:3:7, hwdownload" OUTPUT
RGB colorspace color keying.
The filter accepts the following options:
0.01 matches only the exact key color, while 1.0 matches everything.
0.0 makes pixels either fully transparent, or not transparent at all.
Higher values result in semi-transparent pixels, with a higher transparency the more similar the pixels color is to the key color.
Examples
-i INPUT -vf "hwupload, colorkey_opencl=green:0.3:0.1, hwdownload" OUTPUT
Apply convolution of 3x3, 5x5, 7x7 matrix.
The filter accepts the following options:
Examples
-i INPUT -vf "hwupload, convolution_opencl=0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0, hwdownload" OUTPUT
-i INPUT -vf "hwupload, convolution_opencl=1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1/9:1/9:1/9:1/9, hwdownload" OUTPUT
-i INPUT -vf "hwupload, convolution_opencl=0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:5:1:1:1:0:128:128:128, hwdownload" OUTPUT
-i INPUT -vf "hwupload, convolution_opencl=0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:5:5:5:1:0:128:128:128, hwdownload" OUTPUT
-i INPUT -vf "hwupload, convolution_opencl=1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:1 1 1 1 -8 1 1 1 1:5:5:5:1:0:128:128:0, hwdownload" OUTPUT
-i INPUT -vf "hwupload, convolution_opencl=-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2, hwdownload" OUTPUT
Apply erosion effect to the video.
This filter replaces the pixel by the local(3x3) minimum.
It accepts the following options:
Flags to local 3x3 coordinates region centered on "x":
1 2 3
4 x 5
6 7 8
Example
-i INPUT -vf "hwupload, erosion_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
Feature-point based video stabilization filter.
The filter accepts the following options:
Note that in order to see console debug output you will also need to pass "-v verbose" to ffmpeg.
Viewing point matches in the output video is only supported for RGB input.
Defaults to 0.
Defaults to 1.
This can be turned off for a slight performance gain at the cost of precision.
Defaults to 1.
1.0 is the maximum smoothing strength while values less than that result in less smoothing.
0.0 causes the filter to adaptively choose a smoothing strength on a per-frame basis.
Defaults to 0.0.
The size of the smoothing window is determined by multiplying the framerate of the video by this number.
Acceptable values range from 0.1 to 10.0.
Larger values increase the amount of motion data available for determining how to smooth the camera path, potentially improving smoothness, but also increase latency and memory usage.
Defaults to 2.0.
Examples
-i INPUT -vf "hwupload, deshake_opencl=smooth_strength=0.5, hwdownload" OUTPUT
-i INPUT -filter_complex "[0:v]format=rgba, hwupload, deshake_opencl=debug=1, hwdownload, format=rgba, format=yuv420p" -v verbose OUTPUT
Apply dilation effect to the video.
This filter replaces the pixel by the local(3x3) maximum.
It accepts the following options:
Flags to local 3x3 coordinates region centered on "x":
1 2 3
4 x 5
6 7 8
Example
-i INPUT -vf "hwupload, dilation_opencl=30:40:50:coordinates=231, hwdownload" OUTPUT
Non-local Means denoise filter through OpenCL, this filter accepts same options as nlmeans.
Overlay one video on top of another.
It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid. This filter requires same memory layout for all the inputs. So, format conversion may be needed.
The filter accepts the following options:
Examples
-i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
-i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_opencl, hwdownload" OUTPUT
Add paddings to the input image, and place the original input at the provided x, y coordinates.
It accepts the following options:
The width expression can reference the value set by the height expression, and vice versa.
The default value of width and height is 0.
The x expression can reference the value set by the y expression, and vice versa.
The default value of x and y is 0.
If x or y evaluate to a negative number, they'll be changed so the input image is centered on the padded area.
The value for the width, height, x, and y options are expressions containing the following constants:
Apply the Prewitt operator (<https://en.wikipedia.org/wiki/Prewitt_operator>) to input video stream.
The filter accepts the following option:
Example
-i INPUT -vf "hwupload, prewitt_opencl=scale=2:delta=10, hwdownload" OUTPUT
Filter video using an OpenCL program.
The "program_opencl" filter also supports the framesync options.
The program source file must contain a kernel function with the given name, which will be run once for each plane of the output. Each run on a plane gets enqueued as a separate 2D global NDRange with one work-item for each pixel to be generated. The global ID offset for each work-item is therefore the coordinates of a pixel in the destination image.
The kernel function needs to take the following arguments:
This image will become the output; the kernel should write all of it.
This is a counter starting from zero and increasing by one for each frame.
These are the most recent images on each input. The kernel may read from them to generate the output, but they can't be written to.
Example programs:
__kernel void copy(__write_only image2d_t destination,
unsigned int index,
__read_only image2d_t source)
{
const sampler_t sampler = CLK_NORMALIZED_COORDS_FALSE;
int2 location = (int2)(get_global_id(0), get_global_id(1));
float4 value = read_imagef(source, sampler, location);
write_imagef(destination, location, value);
}
__kernel void rotate_image(__write_only image2d_t dst,
unsigned int index,
__read_only image2d_t src)
{
const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
CLK_FILTER_LINEAR);
float angle = (float)index / 100.0f;
float2 dst_dim = convert_float2(get_image_dim(dst));
float2 src_dim = convert_float2(get_image_dim(src));
float2 dst_cen = dst_dim / 2.0f;
float2 src_cen = src_dim / 2.0f;
int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
float2 dst_pos = convert_float2(dst_loc) - dst_cen;
float2 src_pos = {
cos(angle) * dst_pos.x - sin(angle) * dst_pos.y,
sin(angle) * dst_pos.x + cos(angle) * dst_pos.y
};
src_pos = src_pos * src_dim / dst_dim;
float2 src_loc = src_pos + src_cen;
if (src_loc.x < 0.0f || src_loc.y < 0.0f ||
src_loc.x > src_dim.x || src_loc.y > src_dim.y)
write_imagef(dst, dst_loc, 0.5f);
else
write_imagef(dst, dst_loc, read_imagef(src, sampler, src_loc));
}
__kernel void blend_images(__write_only image2d_t dst,
unsigned int index,
__read_only image2d_t src1,
__read_only image2d_t src2)
{
const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
CLK_FILTER_LINEAR);
float blend = (cos((float)index / 50.0f) + 1.0f) / 2.0f;
int2 dst_loc = (int2)(get_global_id(0), get_global_id(1));
int2 src1_loc = dst_loc * get_image_dim(src1) / get_image_dim(dst);
int2 src2_loc = dst_loc * get_image_dim(src2) / get_image_dim(dst);
float4 val1 = read_imagef(src1, sampler, src1_loc);
float4 val2 = read_imagef(src2, sampler, src2_loc);
write_imagef(dst, dst_loc, val1 * blend + val2 * (1.0f - blend));
}
Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
Destination pixel at position (X, Y) will be picked from source (x, y) position where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero value for pixel will be used for destination pixel.
Xmap and Ymap input video streams must be of same dimensions. Output video stream will have Xmap/Ymap video stream dimensions. Xmap and Ymap input video streams are 32bit float pixel format, single channel.
Apply the Roberts cross operator (<https://en.wikipedia.org/wiki/Roberts_cross>) to input video stream.
The filter accepts the following option:
Example
-i INPUT -vf "hwupload, roberts_opencl=scale=2:delta=10, hwdownload" OUTPUT
Apply the Sobel operator (<https://en.wikipedia.org/wiki/Sobel_operator>) to input video stream.
The filter accepts the following option:
Example
-i INPUT -vf "hwupload, sobel_opencl=scale=2:delta=10, hwdownload" OUTPUT
Perform HDR(PQ/HLG) to SDR conversion with tone-mapping.
It accepts the following parameters:
The default value is 0.5, and the algorithm here is a little different from the cpu version tonemap currently. A setting of 0.0 disables this option.
Currently supported formats are:
Possible values are:
Default is same as input.
Possible values are:
Default is same as input.
Possible values are:
Default is bt709.
Possible value are:
Default is same as input.
Example
-i INPUT -vf "format=p010,hwupload,tonemap_opencl=t=bt2020:tonemap=linear:format=p010,hwdownload,format=p010" OUTPUT
Sharpen or blur the input video.
It accepts the following parameters:
Negative values will blur the input video, while positive values will sharpen it, a value of zero will disable the effect.
Negative values will blur the input video, while positive values will sharpen it, a value of zero will disable the effect.
All parameters are optional and default to the equivalent of the string '5:5:1.0:5:5:0.0'.
Examples
-i INPUT -vf "hwupload, unsharp_opencl=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5, hwdownload" OUTPUT
-i INPUT -vf "hwupload, unsharp_opencl=7:7:-2:7:7:-2, hwdownload" OUTPUT
Cross fade two videos with custom transition effect by using OpenCL.
It accepts the following options:
The program source file must contain a kernel function with the given name, which will be run once for each plane of the output. Each run on a plane gets enqueued as a separate 2D global NDRange with one work-item for each pixel to be generated. The global ID offset for each work-item is therefore the coordinates of a pixel in the destination image.
The kernel function needs to take the following arguments:
This image will become the output; the kernel should write all of it.
These are the most recent images on each input. The kernel may read from them to generate the output, but they can't be written to.
Example programs:
__kernel void blend_images(__write_only image2d_t dst,
__read_only image2d_t src1,
__read_only image2d_t src2,
float progress)
{
const sampler_t sampler = (CLK_NORMALIZED_COORDS_FALSE |
CLK_FILTER_LINEAR);
int2 p = (int2)(get_global_id(0), get_global_id(1));
float2 rp = (float2)(get_global_id(0), get_global_id(1));
float2 dim = (float2)(get_image_dim(src1).x, get_image_dim(src1).y);
rp = rp / dim;
float2 dots = (float2)(20.0, 20.0);
float2 center = (float2)(0,0);
float2 unused;
float4 val1 = read_imagef(src1, sampler, p);
float4 val2 = read_imagef(src2, sampler, p);
bool next = distance(fract(rp * dots, &unused), (float2)(0.5, 0.5)) < (progress / distance(rp, center));
write_imagef(dst, p, next ? val1 : val2);
}
VAAPI Video filters are usually used with VAAPI decoder and VAAPI encoder. Below is a description of VAAPI video filters.
To enable compilation of these filters you need to configure FFmpeg with "--enable-vaapi".
To use vaapi filters, you need to setup the vaapi device correctly. For more information, please read <https://trac.ffmpeg.org/wiki/Hardware/VAAPI>
Overlay one video on the top of another.
It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid.
The filter accepts the following options:
Default value is "0" for both expressions.
Default values are 'overlay_iw' for 'w' and 'overlay_ih*w/overlay_iw' for 'h'.
The expressions can contain the following parameters:
This filter also supports the framesync options.
Examples
-i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuv420p, hwupload[b], [a][b]overlay_vaapi" OUTPUT
-i INPUT -i LOGO -filter_complex "[0:v]hwupload[a], [1:v]format=yuva420p, hwupload[b], [a][b]overlay_vaapi=x=200:y=100:w=400:h=300:alpha=1.0, hwdownload, format=nv12" OUTPUT
Perform HDR(High Dynamic Range) to SDR(Standard Dynamic Range) conversion with tone-mapping. It maps the dynamic range of HDR10 content to the SDR content. It currently only accepts HDR10 as input.
It accepts the following parameters:
Currently supported formats are:
Default is nv12.
Default is same as input.
Default is bt709.
Default is same as input.
Example
tonemap_vaapi=format=p010:t=bt2020-10
Stack input videos horizontally.
This is the VA-API variant of the hstack filter, each input stream may have different height, this filter will scale down/up each input stream while keeping the orignal aspect.
It accepts the following options:
Stack input videos vertically.
This is the VA-API variant of the vstack filter, each input stream may have different width, this filter will scale down/up each input stream while keeping the orignal aspect.
It accepts the following options:
Stack video inputs into custom layout.
This is the VA-API variant of the xstack filter, each input stream may have different size, this filter will scale down/up each input stream to the given output size, or the size of the first input stream.
It accepts the following options:
xstack_vaapi=inputs=4:layout=0_0_1920x1080|0_h0_1920x1080|w0_0_1920x1080|w0_h0_1920x1080
Below is a description of the currently available Vulkan video filters.
To enable compilation of these filters you need to configure FFmpeg with "--enable-vulkan" and either "--enable-libglslang" or "--enable-libshaderc".
Running Vulkan filters requires you to initialize a hardware device and to pass that device to all filters in any filter graph.
For more detailed information see <https://www.ffmpeg.org/ffmpeg.html#Advanced-Video-options>
-init_hw_device vulkan=vk:0 -filter_hw_device vk -i INPUT -vf "hwupload,nlmeans_vulkan,hwdownload" OUTPUT
As Vulkan filters are not able to access frame data in normal memory, all frame data needs to be uploaded (hwupload) to hardware surfaces connected to the appropriate device before being used and then downloaded (hwdownload) back to normal memory. Note that hwupload will upload to a frame with the same layout as the software frame, so it may be necessary to add a format filter immediately before to get the input into the right format and hwdownload does not support all formats on the output - it is usually necessary to insert an additional format filter immediately following in the graph to get the output in a supported format.
Apply an average blur filter, implemented on the GPU using Vulkan.
The filter accepts the following options:
Blend two Vulkan frames into each other.
The "blend" filter takes two input streams and outputs one stream, the first input is the "top" layer and second input is "bottom" layer. By default, the output terminates when the longest input terminates.
A description of the accepted options follows.
Available values for component modes are:
Deinterlacer using bwdif, the "Bob Weaver Deinterlacing Filter" algorithm, implemented on the GPU using Vulkan.
It accepts the following parameters:
The default value is "send_field".
The default value is "auto". If the interlacing is unknown or the decoder does not export this information, top field first will be assumed.
The default value is "all".
Apply an effect that emulates chromatic aberration. Works best with RGB inputs, but provides a similar effect with YCbCr inputs too.
Video source that creates a Vulkan frame of a solid color. Useful for benchmarking, or overlaying.
It accepts the following parameters:
This allows the autodetected value to be overridden as well as allows forcing a specific value used for the output and encoder. If not specified, the range depends on the pixel format. Possible values:
Flips an image vertically.
Flips an image horizontally.
Flips an image along both the vertical and horizontal axis.
Apply Gaussian blur filter on Vulkan frames.
The filter accepts the following options:
Denoise frames using Non-Local Means algorithm, implemented on the GPU using Vulkan. Supports more pixel formats than nlmeans or nlmeans_opencl, including alpha channel support.
The filter accepts the following options.
Overlay one video on top of another.
It takes two inputs and has one output. The first input is the "main" video on which the second input is overlaid. This filter requires all inputs to use the same pixel format. So, format conversion may be needed.
The filter accepts the following options:
Transpose rows with columns in the input video and optionally flip it. For more in depth examples see the transpose video filter, which shares mostly the same options.
It accepts the following parameters:
Can assume the following values:
Transpose rows with columns in the input video and optionally flip it. For more in depth examples see the transpose video filter, which shares mostly the same options.
It accepts the following parameters:
Can assume the following values:
Below is a description of the currently available QSV video filters.
To enable compilation of these filters you need to configure FFmpeg with "--enable-libmfx" or "--enable-libvpl".
To use QSV filters, you need to setup the QSV device correctly. For more information, please read <https://trac.ffmpeg.org/wiki/Hardware/QuickSync>
Stack input videos horizontally.
This is the QSV variant of the hstack filter, each input stream may have different height, this filter will scale down/up each input stream while keeping the orignal aspect.
It accepts the following options:
Stack input videos vertically.
This is the QSV variant of the vstack filter, each input stream may have different width, this filter will scale down/up each input stream while keeping the orignal aspect.
It accepts the following options:
Stack video inputs into custom layout.
This is the QSV variant of the xstack filter.
It accepts the following options:
xstack_qsv=inputs=4:layout=0_0_1920x1080|0_h0_1920x1080|w0_0_1920x1080|w0_h0_1920x1080
Below is a description of the currently available video sources.
Buffer video frames, and make them available to the filter chain.
This source is mainly intended for a programmatic use, in particular through the interface defined in libavfilter/buffersrc.h.
It accepts the following parameters:
For example:
buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
will instruct the source to accept video frames with size 320x240 and with format "yuv410p", assuming 1/24 as the timestamps timebase and square pixels (1:1 sample aspect ratio). Since the pixel format with name "yuv410p" corresponds to the number 6 (check the enum AVPixelFormat definition in libavutil/pixfmt.h), this example corresponds to:
buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
Alternatively, the options can be specified as a flat string, but this syntax is deprecated:
width:height:pix_fmt:time_base.num:time_base.den:pixel_aspect.num:pixel_aspect.den
Create a pattern generated by an elementary cellular automaton.
The initial state of the cellular automaton can be defined through the filename and pattern options. If such options are not specified an initial state is created randomly.
At each new frame a new row in the video is filled with the result of the cellular automaton next generation. The behavior when the whole frame is filled is defined by the scroll option.
This source accepts the following options:
Each non-whitespace character in the string is considered an alive cell, a newline will terminate the row, and further characters in the string will be ignored.
This option is ignored when a file or a pattern is specified.
If filename or pattern is specified, the size is set by default to the width of the specified initial state row, and the height is set to width * PHI.
If size is set, it must contain the width of the specified pattern string, and the specified pattern will be centered in the larger row.
If a filename or a pattern string is not specified, the size value defaults to "320x518" (used for a randomly generated initial state).
Examples
cellauto=f=pattern:s=200x400
cellauto=ratio=2/3:s=200x200
cellauto=p=@s=100x400:full=0:rule=18
cellauto=p='@@ @ @@':s=100x400:full=0:rule=18
Video source generated on GPU using Apple's CoreImage API on OSX.
This video source is a specialized version of the coreimage video filter. Use a core image generator at the beginning of the applied filterchain to generate the content.
The coreimagesrc video source accepts the following options:
list_generators=true
If not specified, or the expressed duration is negative, the video is supposed to be generated forever.
Additionally, all options of the coreimage video filter are accepted. A complete filterchain can be used for further processing of the generated input without CPU-HOST transfer. See coreimage documentation and examples for details.
Examples
ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@inputMessage=https\\\\\://FFmpeg.org/@inputCorrectionLevel=H -frames:v 1 QRCode.png
This example is equivalent to the QRCode example of coreimage without the need for a nullsrc video source.
Captures the Windows Desktop via Desktop Duplication API.
The filter exclusively returns D3D11 Hardware Frames, for on-gpu encoding or processing. So an explicit hwdownload is needed for any kind of software processing.
It accepts the following options:
Usually corresponds to the index Windows has given the screen minus one, so it's starting at 0.
Defaults to output 0.
Defaults to true.
Only affects hardware cursors. If a game or application renders its own cursor, it'll always be captured.
Defaults to 30 FPS.
Defaults to the full size of the screen.
Cropped from the bottom/right if smaller than screen size.
It accepts the following values:
Examples
Capture primary screen and encode using nvenc:
ffmpeg -f lavfi -i ddagrab -c:v h264_nvenc -cq 18 output.mp4
You can also skip the lavfi device and directly use the filter. Also demonstrates downloading the frame and encoding with libx264. Explicit output format specification is required in this case:
ffmpeg -filter_complex ddagrab=output_idx=1:framerate=60,hwdownload,format=bgra -c:v libx264 -crf 18 output.mp4
If you want to capture only a subsection of the desktop, this can be achieved by specifying a smaller size and its offsets into the screen:
ddagrab=video_size=800x600:offset_x=100:offset_y=100
Generate several gradients.
If not specified, or the expressed duration is negative, the video is supposed to be generated forever.
Generate a Mandelbrot set fractal, and progressively zoom towards the point specified with start_x and start_y.
This source accepts the following options:
It shall assume one of the following values:
Default value is mincol.
Default value is normalized_iteration_count.
Generate various test patterns, as generated by the MPlayer test filter.
The size of the generated video is fixed, and is 256x256. This source is useful in particular for testing encoding features.
This source accepts the following options:
If not specified, or the expressed duration is negative, the video is supposed to be generated forever.
Default value is "all", which will cycle through the list of all tests.
Some examples:
mptestsrc=t=dc_luma
will generate a "dc_luma" test pattern.
Provide a frei0r source.
To enable compilation of this filter you need to install the frei0r header and configure FFmpeg with "--enable-frei0r".
This source accepts the following parameters:
For example, to generate a frei0r partik0l source with size 200x200 and frame rate 10 which is overlaid on the overlay filter main input:
frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
Generate a life pattern.
This source is based on a generalization of John Conway's life game.
The sourced input represents a life grid, each pixel represents a cell which can be in one of two possible states, alive or dead. Every cell interacts with its eight neighbours, which are the cells that are horizontally, vertically, or diagonally adjacent.
At each interaction the grid evolves according to the adopted rule, which specifies the number of neighbor alive cells which will make a cell stay alive or born. The rule option allows one to specify the rule to adopt.
This source accepts the following options:
If this option is not specified, the initial grid is generated randomly.
A rule can be specified with a code of the kind "SNS/BNB", where NS and NB are sequences of numbers in the range 0-8, NS specifies the number of alive neighbor cells which make a live cell stay alive, and NB the number of alive neighbor cells which make a dead cell to become alive (i.e. to "born"). "s" and "b" can be used in place of "S" and "B", respectively.
Alternatively a rule can be specified by an 18-bits integer. The 9 high order bits are used to encode the next cell state if it is alive for each number of neighbor alive cells, the low order bits specify the rule for "borning" new cells. Higher order bits encode for an higher number of neighbor cells. For example the number 6153 = "(12<<9)+9" specifies a stay alive rule of 12 and a born rule of 9, which corresponds to "S23/B03".
Default value is "S23/B3", which is the original Conway's game of life rule, and will keep a cell alive if it has 2 or 3 neighbor alive cells, and will born a new cell if there are three alive cells around a dead cell.
If filename is specified, the size is set by default to the same size of the input file. If size is set, it must contain the size specified in the input file, and the initial grid defined in that file is centered in the larger resulting area.
If a filename is not specified, the size value defaults to "320x240" (used for a randomly generated initial grid).
For the syntax of these 3 color options, check the "Color" section in the ffmpeg-utils manual.
Examples
life=f=pattern:s=300x300
life=ratio=2/3:s=200x200
life=rule=S14/B34
ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
The "allrgb" source returns frames of size 4096x4096 of all rgb colors.
The "allyuv" source returns frames of size 4096x4096 of all yuv colors.
The "color" source provides an uniformly colored input.
The "colorchart" source provides a colors checker chart.
The "colorspectrum" source provides a color spectrum input.
The "haldclutsrc" source provides an identity Hald CLUT. See also haldclut filter.
The "nullsrc" source returns unprocessed video frames. It is mainly useful to be employed in analysis / debugging tools, or as the source for filters which ignore the input data.
The "pal75bars" source generates a color bars pattern, based on EBU PAL recommendations with 75% color levels.
The "pal100bars" source generates a color bars pattern, based on EBU PAL recommendations with 100% color levels.
The "rgbtestsrc" source generates an RGB test pattern useful for detecting RGB vs BGR issues. You should see a red, green and blue stripe from top to bottom.
The "smptebars" source generates a color bars pattern, based on the SMPTE Engineering Guideline EG 1-1990.
The "smptehdbars" source generates a color bars pattern, based on the SMPTE RP 219-2002.
The "testsrc" source generates a test video pattern, showing a color pattern, a scrolling gradient and a timestamp. This is mainly intended for testing purposes.
The "testsrc2" source is similar to testsrc, but supports more pixel formats instead of just "rgb24". This allows using it as an input for other tests without requiring a format conversion.
The "yuvtestsrc" source generates an YUV test pattern. You should see a y, cb and cr stripe from top to bottom.
The sources accept the following parameters:
This option is not available with the "allrgb", "allyuv", and "haldclutsrc" filters.
If not specified, or the expressed duration is negative, the video is supposed to be generated forever.
Since the frame rate is used as time base, all frames including the last one will have their full duration. If the specified duration is not a multiple of the frame duration, it will be rounded up.
The displayed timestamp value will correspond to the original timestamp value multiplied by the power of 10 of the specified value. Default value is 0.
Available values are:
Default value is "reference".
Examples
testsrc=duration=5.3:size=qcif:rate=10
color=c=red@0.2:s=qcif:r=10
nullsrc=s=256x256, geq=random(1)*255:128:128
Commands
The "color" source supports the following commands:
Generate video using an OpenCL program.
For details of how the program loading works, see the program_opencl filter.
Example programs:
__kernel void ramp(__write_only image2d_t dst,
unsigned int index)
{
int2 loc = (int2)(get_global_id(0), get_global_id(1));
float4 val;
val.xy = val.zw = convert_float2(loc) / convert_float2(get_image_dim(dst));
write_imagef(dst, loc, val);
}
__kernel void sierpinski_carpet(__write_only image2d_t dst,
unsigned int index)
{
int2 loc = (int2)(get_global_id(0), get_global_id(1));
float4 value = 0.0f;
int x = loc.x + index;
int y = loc.y + index;
while (x > 0 || y > 0) {
if (x % 3 == 1 && y % 3 == 1) {
value = 1.0f;
break;
}
x /= 3;
y /= 3;
}
write_imagef(dst, loc, value);
}
Generate a Sierpinski carpet/triangle fractal, and randomly pan around.
This source accepts the following options:
Generate a zoneplate test video pattern.
This source accepts the following options:
If not specified, or the expressed duration is negative, the video is supposed to be generated forever.
Commands
This source supports the some above options as commands.
Examples
zoneplate=ku=512:kv=0:kt2=0:kx2=256:s=wvga:xo=-426:kt=11
zoneplate=ku=512:kv=0:kt2=0:ky2=156:s=wvga:yo=-240:kt=11
zoneplate=ku=512:kv=100:kt2=0:ky2=256:kx2=556:s=wvga:yo=0:kt=11
Below is a description of the currently available video sinks.
Buffer video frames, and make them available to the end of the filter graph.
This sink is mainly intended for programmatic use, in particular through the interface defined in libavfilter/buffersink.h or the options system.
It accepts a pointer to an AVBufferSinkContext structure, which defines the incoming buffers' formats, to be passed as the opaque parameter to "avfilter_init_filter" for initialization.
Null video sink: do absolutely nothing with the input video. It is mainly useful as a template and for use in analysis / debugging tools.
Below is a description of the currently available multimedia filters.
Convert input audio to 3d scope video output.
The filter accepts the following options:
Commands
Filter supports the some above options as commands.
Convert input audio to a video output, displaying the audio bit scope.
The filter accepts the following options:
Draw a graph using input audio metadata.
See drawgraph
See graphmonitor.
Convert input audio to a video output, displaying the volume histogram.
The filter accepts the following options:
It accepts the following values:
Default is "single".
It accepts the following values:
Default is "log".
It accepts the following values:
Default is "log".
It accepts the following values:
Default is "replace".
It accepts the following values:
Default is "abs".
Measures phase of input audio, which is exported as metadata "lavfi.aphasemeter.phase", representing mean phase of current audio frame. A video output can also be produced and is enabled by default. The audio is passed through as first output.
Audio will be rematrixed to stereo if it has a different channel layout. Phase value is in range "[-1, 1]" where -1 means left and right channels are completely out of phase and 1 means channels are in phase.
The filter accepts the following options, all related to its video output:
phasing detection
The filter also detects out of phase and mono sequences in stereo streams. It logs the sequence start, end and duration when it lasts longer or as long as the minimum set.
The filter accepts the following options for this detection:
Examples
ffmpeg -i stereo.wav -af aphasemeter=video=0:phasing=1:duration=1:tolerance=0.001 -f null -
Convert input audio to a video output, representing the audio vector scope.
The filter is used to measure the difference between channels of stereo audio stream. A monaural signal, consisting of identical left and right signal, results in straight vertical line. Any stereo separation is visible as a deviation from this line, creating a Lissajous figure. If the straight (or deviation from it) but horizontal line appears this indicates that the left and right channels are out of phase.
The filter accepts the following options:
Available values are:
Default value is lissajous.
Available values are:
Default value is dot.
Available values are:
Examples
ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
[a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
Commands
This filter supports the all above options as commands except options "size" and "rate".
Benchmark part of a filtergraph.
The filter accepts the following options:
Available values are:
Examples
bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
Concatenate audio and video streams, joining them together one after the other.
The filter works on segments of synchronized video and audio streams. All segments must have the same number of streams of each type, and that will also be the number of streams at output.
The filter accepts the following options:
The filter has v+a outputs: first v video outputs, then a audio outputs.
There are nx(v+a) inputs: first the inputs for the first segment, in the same order as the outputs, then the inputs for the second segment, etc.
Related streams do not always have exactly the same duration, for various reasons including codec frame size or sloppy authoring. For that reason, related synchronized streams (e.g. a video and its audio track) should be concatenated at once. The concat filter will use the duration of the longest stream in each segment (except the last one), and if necessary pad shorter audio streams with silence.
For this filter to work correctly, all segments must start at timestamp 0.
All corresponding streams must have the same parameters in all segments; the filtering system will automatically select a common pixel format for video streams, and a common sample format, sample rate and channel layout for audio streams, but other settings, such as resolution, must be converted explicitly by the user.
Different frame rates are acceptable but will result in variable frame rate at output; be sure to configure the output file to handle it.
Examples
ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
'[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
concat=n=3:v=1:a=2 [v] [a1] [a2]' \
-map '[v]' -map '[a1]' -map '[a2]' output.mkv
movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
[v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
Note that a desync will happen at the stitch if the audio and video streams do not have exactly the same duration in the first file.
Commands
This filter supports the following commands:
EBU R128 scanner filter. This filter takes an audio stream and analyzes its loudness level. By default, it logs a message at a frequency of 10Hz with the Momentary loudness (identified by "M"), Short-term loudness ("S"), Integrated loudness ("I") and Loudness Range ("LRA").
The filter can only analyze streams which have sample format is double-precision floating point. The input stream will be converted to this specification, if needed. Users may need to insert aformat and/or aresample filters after this filter to obtain the original parameters.
The filter also has a video output (see the video option) with a real time graph to observe the loudness evolution. The graphic contains the logged message mentioned above, so it is not printed anymore when this option is set, unless the verbose logging is set. The main graphing area contains the short-term loudness (3 seconds of analysis), and the gauge on the right is for the momentary loudness (400 milliseconds), but can optionally be configured to instead display short-term loudness (see gauge).
The green area marks a +/- 1LU target range around the target loudness (-23LUFS by default, unless modified through target).
More information about the Loudness Recommendation EBU R128 on <http://tech.ebu.ch/loudness>.
The filter accepts the following options:
Default is 0.
Available values are:
By default, the logging level is set to info. If the video or the metadata options are set, it switches to verbose.
Available modes can be cumulated (the option is a "flag" type). Possible values are:
Simple peak mode looking for the higher sample value. It logs a message for sample-peak (identified by "SPK").
If enabled, the peak lookup is done on an over-sampled version of the input stream for better peak accuracy. It logs a message for true-peak. (identified by "TPK") and true-peak per frame (identified by "FTPK"). This mode requires a build with "libswresample".
Examples
ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
Temporally interleave frames from several inputs.
"interleave" works with video inputs, "ainterleave" with audio.
These filters read frames from several inputs and send the oldest queued frame to the output.
Input streams must have well defined, monotonically increasing frame timestamp values.
In order to submit one frame to output, these filters need to enqueue at least one frame for each input, so they cannot work in case one input is not yet terminated and will not receive incoming frames.
For example consider the case when one input is a "select" filter which always drops input frames. The "interleave" filter will keep reading from that input, but it will never be able to send new frames to output until the input sends an end-of-stream signal.
Also, depending on inputs synchronization, the filters will drop frames in case one input receives more frames than the other ones, and the queue is already filled.
These filters accept the following options:
Examples
ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
Measure filtering latency.
Report previous filter filtering latency, delay in number of audio samples for audio filters or number of video frames for video filters.
On end of input stream, filter will report min and max measured latency for previous running filter in filtergraph.
Manipulate frame metadata.
This filter accepts the following options:
Can be one of the following:
Can be one of following:
Examples
signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
silencedetect,ametadata=mode=print:file=metadata.txt
metadata=mode=print:file='pipe\:4'
Set read/write permissions for the output frames.
These filters are mainly aimed at developers to test direct path in the following filter in the filtergraph.
The filters accept the following options:
It accepts the following values:
Note: in case of auto-inserted filter between the permission filter and the following one, the permission might not be received as expected in that following filter. Inserting a format or aformat filter before the perms/aperms filter can avoid this problem.
Slow down filtering to match real time approximately.
These filters will pause the filtering for a variable amount of time to match the output rate with the input timestamps. They are similar to the re option to "ffmpeg".
They accept the following options:
A processing speed faster than what is possible without these filters cannot be achieved.
Commands
Both filters supports the all above options as commands.
Split single input stream into multiple streams.
This filter does opposite of concat filters.
"segment" works on video frames, "asegment" on audio samples.
This filter accepts the following options:
In all cases, prefixing an each segment with '+' will make it relative to the previous segment.
Examples
asegment=timestamps="60|150"
Select frames to pass in output.
This filter accepts the following options:
If the expression is evaluated to zero, the frame is discarded.
If the evaluation result is negative or NaN, the frame is sent to the first output; otherwise it is sent to the output with index "ceil(val)-1", assuming that the input index starts from 0.
For example a value of 1.2 corresponds to the output with index "ceil(1.2)-1 = 2-1 = 1", that is the second output.
The expression can contain the following constants:
This works by comparing the frame pts against the lavf.concat.start_time and the lavf.concat.duration packet metadata values which are also present in the decoded frames.
The concatdec_select variable is -1 if the frame pts is at least start_time and either the duration metadata is missing or the frame pts is less than start_time + duration, 0 otherwise, and NaN if the start_time metadata is missing.
That basically means that an input frame is selected if its pts is within the interval set by the concat demuxer.
The default value of the select expression is "1".
Examples
select
The example above is the same as:
select=1
select=0
select='eq(pict_type\,I)'
select='not(mod(n\,100))'
select=between(t\,10\,20)
select=between(t\,10\,20)*eq(pict_type\,I)
select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
aselect='gt(samples_n\,100)'
ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
Comparing scene against a value between 0.3 and 0.5 is generally a sane choice.
select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
Send commands to filters in the filtergraph.
These filters read commands to be sent to other filters in the filtergraph.
"sendcmd" must be inserted between two video filters, "asendcmd" must be inserted between two audio filters, but apart from that they act the same way.
The specification of commands can be provided in the filter arguments with the commands option, or in a file specified by the filename option.
These filters accept the following options:
Commands syntax
A commands description consists of a sequence of interval specifications, comprising a list of commands to be executed when a particular event related to that interval occurs. The occurring event is typically the current frame time entering or leaving a given time interval.
An interval is specified by the following syntax:
<START>[-<END>] <COMMANDS>;
The time interval is specified by the START and END times. END is optional and defaults to the maximum time.
The current frame time is considered within the specified interval if it is included in the interval [START, END), that is when the time is greater or equal to START and is lesser than END.
COMMANDS consists of a sequence of one or more command specifications, separated by ",", relating to that interval. The syntax of a command specification is given by:
[<FLAGS>] <TARGET> <COMMAND> <ARG>
FLAGS is optional and specifies the type of events relating to the time interval which enable sending the specified command, and must be a non-null sequence of identifier flags separated by "+" or "|" and enclosed between "[" and "]".
The following flags are recognized:
The expression is evaluated through the eval API and can contain the following constants:
If FLAGS is not specified, a default value of "[enter]" is assumed.
TARGET specifies the target of the command, usually the name of the filter class or a specific filter instance name.
COMMAND specifies the name of the command for the target filter.
ARG is optional and specifies the optional list of argument for the given COMMAND.
Between one interval specification and another, whitespaces, or sequences of characters starting with "#" until the end of line, are ignored and can be used to annotate comments.
A simplified BNF description of the commands specification syntax follows:
<COMMAND_FLAG> ::= "enter" | "leave"
<COMMAND_FLAGS> ::= <COMMAND_FLAG> [(+|"|")<COMMAND_FLAG>]
<COMMAND> ::= ["[" <COMMAND_FLAGS> "]"] <TARGET> <COMMAND> [<ARG>]
<COMMANDS> ::= <COMMAND> [,<COMMANDS>]
<INTERVAL> ::= <START>[-<END>] <COMMANDS>
<INTERVALS> ::= <INTERVAL>[;<INTERVALS>]
Examples
asendcmd=c='4.0 atempo tempo 1.5',atempo
asendcmd=c='4.0 atempo@my tempo 1.5',atempo@my
# show text in the interval 5-10
5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
[leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
# desaturate the image in the interval 15-20
15.0-20.0 [enter] hue s 0,
[enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
[leave] hue s 1,
[leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
# apply an exponential saturation fade-out effect, starting from time 25
25 [enter] hue s exp(25-t)
A filtergraph allowing to read and process the above command list stored in a file test.cmd, can be specified with:
sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
Change the PTS (presentation timestamp) of the input frames.
"setpts" works on video frames, "asetpts" on audio frames.
This filter accepts the following options:
The expression is evaluated through the eval API and can contain the following constants:
Examples
setpts=PTS-STARTPTS
setpts=0.5*PTS
setpts=2.0*PTS
setpts=N/(25*TB)
setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
setpts=PTS+10/TB
setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
asetpts=N/SR/TB
Commands
Both filters support all above options as commands.
Force color range for the output video frame.
The "setrange" filter marks the color range property for the output frames. It does not change the input frame, but only sets the corresponding property, which affects how the frame is treated by following filters.
The filter accepts the following options:
Set the timebase to use for the output frames timestamps. It is mainly useful for testing timebase configuration.
It accepts the following parameters:
The value for tb is an arithmetic expression representing a rational. The expression can contain the constants "AVTB" (the default timebase), "intb" (the input timebase) and "sr" (the sample rate, audio only). Default value is "intb".
Examples
settb=expr=1/25
settb=expr=0.1
settb=1+0.001
settb=2*intb
settb=AVTB
Convert input audio to a video output representing frequency spectrum logarithmically using Brown-Puckette constant Q transform algorithm with direct frequency domain coefficient calculation (but the transform itself is not really constant Q, instead the Q factor is actually variable/clamped), with musical tone scale, from E0 to D#10.
The filter accepts the following options:
and functions:
Default value is 16.
and functions:
Default value is "sono_v".
Default value is "384*tc/(384+tc*f)".
and functions:
Default value is "st(0, (midi(f)-59.5)/12); st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0)); r(1-ld(1)) + b(ld(1))".
Examples
ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
sono_h=0
ffplay -f lavfi 'aevalsrc=0.1*sin(2*PI*55*t)+0.1*sin(4*PI*55*t)+0.1*sin(6*PI*55*t)+0.1*sin(8*PI*55*t),
asplit[a][out1]; [a] showcqt [out0]'
ffplay -f lavfi 'aevalsrc=0.1*sin(2*PI*55*t)+0.1*sin(4*PI*55*t)+0.1*sin(6*PI*55*t)+0.1*sin(8*PI*55*t),
asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
bar_v=10:sono_v=bar_v*a_weighting(f)
bar_g=2:sono_g=2
tc=0.33:tlength='st(0,0.17); 384*tc / (384 / ld(0) + tc*f /(1-ld(0))) + 384*tc / (tc*f / ld(0) + 384 /(1-ld(0)))'
fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
font='Courier New,Monospace,mono|bold'
axisfile=myaxis.png:basefreq=40:endfreq=10000
Convert input audio to video output representing frequency spectrum using Continuous Wavelet Transform and Morlet wavelet.
The filter accepts the following options:
Default value is "linear".
Default value is "log".
Default value is "magnitude".
Convert input audio to video output representing the audio power spectrum. Audio amplitude is on Y-axis while frequency is on X-axis.
The filter accepts the following options:
It accepts the following values:
Default is "bar".
It accepts the following values:
Default is "log".
It accepts the following values:
Default is "lin".
Default is 2048
It accepts the following values:
Default is "hanning".
It accepts the following values:
Default is "combined".
It accepts the following values:
Default is "magnitude".
Convert stereo input audio to a video output, representing the spatial relationship between two channels.
The filter accepts the following options:
It accepts the following values:
Default value is "hann".
Convert input audio to a video output, representing the audio frequency spectrum.
The filter accepts the following options:
It accepts the following values:
Default value is "replace".
It accepts the following values:
Default value is combined.
It accepts the following values:
Default value is channel.
It accepts the following values:
Default value is sqrt.
It accepts the following values:
Default value is lin.
It accepts the following values:
Default value is "hann".
The usage is very similar to the showwaves filter; see the examples in that section.
Examples
showspectrum=s=1280x480:scale=log
ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
[a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
Convert input audio to a single video frame, representing the audio frequency spectrum.
The filter accepts the following options:
It accepts the following values:
Default value is combined.
It accepts the following values:
Default value is intensity.
It accepts the following values:
Default value is log.
It accepts the following values:
Default value is lin.
It accepts the following values:
Default value is "hann".
Examples
ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
Convert input audio volume to a video output.
The filter accepts the following options:
The expression can use the following variables:
Convert input audio to a video output, representing the samples waves.
The filter accepts the following options:
Available values are:
Default value is "point".
Available values are:
Default is linear.
Available values are:
Default value is "scale".
Examples
amovie=a.mp3,asplit[out0],showwaves[out1]
aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
Convert input audio to a single video frame, representing the samples waves.
The filter accepts the following options:
Available values are:
Default is linear.
Available values are:
Default value is "scale".
Available values are:
Default value is "average".
Examples
ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
Delete frame side data, or select frames based on it.
This filter accepts the following options:
Can be one of the following:
Synthesize audio from 2 input video spectrums, first input stream represents magnitude across time and second represents phase across time. The filter will transform from frequency domain as displayed in videos back to time domain as presented in audio output.
This filter is primarily created for reversing processed showspectrum filter outputs, but can synthesize sound from other spectrograms too. But in such case results are going to be poor if the phase data is not available, because in such cases phase data need to be recreated, usually it's just recreated from random noise. For best results use gray only output ("channel" color mode in showspectrum filter) and "log" scale for magnitude video and "lin" scale for phase video. To produce phase, for 2nd video, use "data" option. Inputs videos should generally use "fullframe" slide mode as that saves resources needed for decoding video.
The filter accepts the following options:
Examples
ffmpeg -i input.flac -lavfi showspectrum=mode=separate:scale=log:overlap=0.875:color=channel:slide=fullframe:data=magnitude -an -c:v rawvideo magnitude.nut
ffmpeg -i input.flac -lavfi showspectrum=mode=separate:scale=lin:overlap=0.875:color=channel:slide=fullframe:data=phase -an -c:v rawvideo phase.nut
ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
Split input into several identical outputs.
"asplit" works with audio input, "split" with video.
The filter accepts a single parameter which specifies the number of outputs. If unspecified, it defaults to 2.
Examples
[in] split [out0][out1]
[in] asplit=3 [out0][out1][out2]
[in] split [splitout1][splitout2];
[splitout1] crop=100:100:0:0 [cropout];
[splitout2] pad=200:200:100:100 [padout];
ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
Receive commands sent through a libzmq client, and forward them to filters in the filtergraph.
"zmq" and "azmq" work as a pass-through filters. "zmq" must be inserted between two video filters, "azmq" between two audio filters. Both are capable to send messages to any filter type.
To enable these filters you need to install the libzmq library and headers and configure FFmpeg with "--enable-libzmq".
For more information about libzmq see: <http://www.zeromq.org/>
The "zmq" and "azmq" filters work as a libzmq server, which receives messages sent through a network interface defined by the bind_address (or the abbreviation "b") option. Default value of this option is tcp://localhost:5555. You may want to alter this value to your needs, but do not forget to escape any ':' signs (see filtergraph escaping).
The received message must be in the form:
<TARGET> <COMMAND> [<ARG>]
TARGET specifies the target of the command, usually the name of the filter class or a specific filter instance name. The default filter instance name uses the pattern Parsed_<filter_name>_<index>, but you can override this by using the filter_name@id syntax (see Filtergraph syntax).
COMMAND specifies the name of the command for the target filter.
ARG is optional and specifies the optional argument list for the given COMMAND.
Upon reception, the message is processed and the corresponding command is injected into the filtergraph. Depending on the result, the filter will send a reply to the client, adopting the format:
<ERROR_CODE> <ERROR_REASON>
<MESSAGE>
MESSAGE is optional.
Examples
Look at tools/zmqsend for an example of a zmq client which can be used to send commands processed by these filters.
Consider the following filtergraph generated by ffplay. In this example the last overlay filter has an instance name. All other filters will have default instance names.
ffplay -dumpgraph 1 -f lavfi "
color=s=100x100:c=red [l];
color=s=100x100:c=blue [r];
nullsrc=s=200x100, zmq [bg];
[bg][l] overlay [bg+l];
[bg+l][r] overlay@my=x=100 "
To change the color of the left side of the video, the following command can be used:
echo Parsed_color_0 c yellow | tools/zmqsend
To change the right side:
echo Parsed_color_1 c pink | tools/zmqsend
To change the position of the right side:
echo overlay@my x 150 | tools/zmqsend
Below is a description of the currently available multimedia sources.
This is the same as movie source, except it selects an audio stream by default.
Generate an Audio/Video Sync Test.
Generated stream periodically shows flash video frame and emits beep in audio. Useful to inspect A/V sync issues.
It accepts the following options:
Commands
This source supports the some above options as commands.
Read audio and/or video stream(s) from a movie container.
It accepts the following parameters:
Note that when the movie is looped the source timestamps are not changed, so it will generate non monotonically increasing timestamps.
ffplay -f lavfi
"movie=filename='1.sdp':format_opts='protocol_whitelist=file,rtp,udp\:protocol_blacklist=http'"
It allows overlaying a second video on top of the main input of a filtergraph, as shown in this graph:
input -----------> deltapts0 --> overlay --> output
^
|
movie --> scale--> deltapts1 -------+
Examples
movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
[in] setpts=PTS-STARTPTS [main];
[main][over] overlay=16:16 [out]
movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
[in] setpts=PTS-STARTPTS [main];
[main][over] overlay=16:16 [out]
movie=dvd.vob:s=v:0+#0x81 [video] [audio]
Commands
Both movie and amovie support the following commands:
FFmpeg can be hooked up with a number of external libraries to add support for more formats. None of them are used by default, their use has to be explicitly requested by passing the appropriate flags to ./configure.
FFmpeg can make use of the AOM library for AV1 decoding and encoding.
Go to <http://aomedia.org/> and follow the instructions for installing the library. Then pass "--enable-libaom" to configure to enable it.
FFmpeg can use the AMD Advanced Media Framework library for accelerated H.264 and HEVC(only windows) encoding on hardware with Video Coding Engine (VCE).
To enable support you must obtain the AMF framework header files(version 1.4.9+) from <https://github.com/GPUOpen-LibrariesAndSDKs/AMF.git>.
Create an "AMF/" directory in the system include path. Copy the contents of "AMF/amf/public/include/" into that directory. Then configure FFmpeg with "--enable-amf".
Initialization of amf encoder occurs in this order: 1) trying to initialize through dx11(only windows) 2) trying to initialize through dx9(only windows) 3) trying to initialize through vulkan
To use h.264(AMD VCE) encoder on linux amdgru-pro version 19.20+ and amf-amdgpu-pro package(amdgru-pro contains, but does not install automatically) are required.
This driver can be installed using amdgpu-pro-install script in official amd driver archive.
FFmpeg can read AviSynth scripts as input. To enable support, pass "--enable-avisynth" to configure after installing the headers provided by <https://github.com/AviSynth/AviSynthPlus>. AviSynth+ can be configured to install only the headers by either passing "-DHEADERS_ONLY:bool=on" to the normal CMake-based build system, or by using the supplied "GNUmakefile".
For Windows, supported AviSynth variants are <http://avisynth.nl> for 32-bit builds and <http://avisynth.nl/index.php/AviSynth+> for 32-bit and 64-bit builds.
For Linux, macOS, and BSD, the only supported AviSynth variant is <https://github.com/AviSynth/AviSynthPlus>, starting with version 3.5.
By default, FFmpeg assumes compatibility with 32-bit MSVC builds of AviSynth+ since that is the most widely-used and entrenched build configuration. Users can override this and enable support for 32-bit GCC builds of AviSynth+ by passing "-DAVSC_WIN32_GCC32" to "--extra-cflags" when configuring FFmpeg.
64-bit builds of FFmpeg are not affected, and can use either MSVC or GCC builds of AviSynth+ without any special flags.
FFmpeg can make use of the Chromaprint library for generating audio fingerprints. Pass "--enable-chromaprint" to configure to enable it. See <https://acoustid.org/chromaprint>.
FFmpeg can make use of the codec2 library for codec2 decoding and encoding. There is currently no native decoder, so libcodec2 must be used for decoding.
Go to <http://freedv.org/>, download "Codec 2 source archive". Build and install using CMake. Debian users can install the libcodec2-dev package instead. Once libcodec2 is installed you can pass "--enable-libcodec2" to configure to enable it.
The easiest way to use codec2 is with .c2 files, since they contain the mode information required for decoding. To encode such a file, use a .c2 file extension and give the libcodec2 encoder the -mode option: "ffmpeg -i input.wav -mode 700C output.c2". Playback is as simple as "ffplay output.c2". For a list of supported modes, run "ffmpeg -h encoder=libcodec2". Raw codec2 files are also supported. To make sense of them the mode in use needs to be specified as a format option: "ffmpeg -f codec2raw -mode 1300 -i input.raw output.wav".
FFmpeg can make use of the dav1d library for AV1 video decoding.
Go to <https://code.videolan.org/videolan/dav1d> and follow the instructions for installing the library. Then pass "--enable-libdav1d" to configure to enable it.
FFmpeg can make use of the davs2 library for AVS2-P2/IEEE1857.4 video decoding.
Go to <https://github.com/pkuvcl/davs2> and follow the instructions for installing the library. Then pass "--enable-libdavs2" to configure to enable it.
FFmpeg can make use of the uavs3d library for AVS3-P2/IEEE1857.10 video decoding.
Go to <https://github.com/uavs3/uavs3d> and follow the instructions for installing the library. Then pass "--enable-libuavs3d" to configure to enable it.
FFmpeg can make use of the Game Music Emu library to read audio from supported video game music file formats. Pass "--enable-libgme" to configure to enable it. See <https://bitbucket.org/mpyne/game-music-emu/overview>.
FFmpeg can use Intel QuickSync Video (QSV) for accelerated decoding and encoding of multiple codecs. To use QSV, FFmpeg must be linked against the "libmfx" dispatcher, which loads the actual decoding libraries.
The dispatcher is open source and can be downloaded from <https://github.com/lu-zero/mfx_dispatch.git>. FFmpeg needs to be configured with the "--enable-libmfx" option and "pkg-config" needs to be able to locate the dispatcher's ".pc" files.
FFmpeg can make use of the Kvazaar library for HEVC encoding.
Go to <https://github.com/ultravideo/kvazaar> and follow the instructions for installing the library. Then pass "--enable-libkvazaar" to configure to enable it.
FFmpeg can make use of the LAME library for MP3 encoding.
Go to <http://lame.sourceforge.net/> and follow the instructions for installing the library. Then pass "--enable-libmp3lame" to configure to enable it.
iLBC is a narrowband speech codec that has been made freely available by Google as part of the WebRTC project. libilbc is a packaging friendly copy of the iLBC codec. FFmpeg can make use of the libilbc library for iLBC decoding and encoding.
Go to <https://github.com/TimothyGu/libilbc> and follow the instructions for installing the library. Then pass "--enable-libilbc" to configure to enable it.
JPEG XL is an image format intended to fully replace legacy JPEG for an extended period of life. See <https://jpegxl.info/> for more information, and see <https://github.com/libjxl/libjxl> for the library source. You can pass "--enable-libjxl" to configure in order enable the libjxl wrapper.
FFmpeg can make use of the libvpx library for VP8/VP9 decoding and encoding.
Go to <http://www.webmproject.org/> and follow the instructions for installing the library. Then pass "--enable-libvpx" to configure to enable it.
FFmpeg can make use of this library, originating in Modplug-XMMS, to read from MOD-like music files. See <https://github.com/Konstanty/libmodplug>. Pass "--enable-libmodplug" to configure to enable it.
Spun off Google Android sources, OpenCore, VisualOn and Fraunhofer libraries provide encoders for a number of audio codecs.
The license of the Fraunhofer AAC library is incompatible with the GPL. Therefore, for GPL builds, you have to pass "--enable-nonfree" to configure in order to use it. To the best of our knowledge, it is compatible with the LGPL.
OpenCORE AMR
FFmpeg can make use of the OpenCORE libraries for AMR-NB decoding/encoding and AMR-WB decoding.
Go to <http://sourceforge.net/projects/opencore-amr/> and follow the instructions for installing the libraries. Then pass "--enable-libopencore-amrnb" and/or "--enable-libopencore-amrwb" to configure to enable them.
VisualOn AMR-WB encoder library
FFmpeg can make use of the VisualOn AMR-WBenc library for AMR-WB encoding.
Go to <http://sourceforge.net/projects/opencore-amr/> and follow the instructions for installing the library. Then pass "--enable-libvo-amrwbenc" to configure to enable it.
Fraunhofer AAC library
FFmpeg can make use of the Fraunhofer AAC library for AAC decoding & encoding.
Go to <http://sourceforge.net/projects/opencore-amr/> and follow the instructions for installing the library. Then pass "--enable-libfdk-aac" to configure to enable it.
FFmpeg can make use of the OpenH264 library for H.264 decoding and encoding.
Go to <http://www.openh264.org/> and follow the instructions for installing the library. Then pass "--enable-libopenh264" to configure to enable it.
For decoding, this library is much more limited than the built-in decoder in libavcodec; currently, this library lacks support for decoding B-frames and some other main/high profile features. (It currently only supports constrained baseline profile and CABAC.) Using it is mostly useful for testing and for taking advantage of Cisco's patent portfolio license (<http://www.openh264.org/BINARY_LICENSE.txt>).
FFmpeg can use the OpenJPEG libraries for decoding/encoding J2K videos. Go to <http://www.openjpeg.org/> to get the libraries and follow the installation instructions. To enable using OpenJPEG in FFmpeg, pass "--enable-libopenjpeg" to ./configure.
FFmpeg can make use of rav1e (Rust AV1 Encoder) via its C bindings to encode videos. Go to <https://github.com/xiph/rav1e/> and follow the instructions to build the C library. To enable using rav1e in FFmpeg, pass "--enable-librav1e" to ./configure.
FFmpeg can make use of the Scalable Video Technology for AV1 library for AV1 encoding.
Go to <https://gitlab.com/AOMediaCodec/SVT-AV1/> and follow the instructions for installing the library. Then pass "--enable-libsvtav1" to configure to enable it.
FFmpeg can make use of the TwoLAME library for MP2 encoding.
Go to <http://www.twolame.org/> and follow the instructions for installing the library. Then pass "--enable-libtwolame" to configure to enable it.
FFmpeg can read VapourSynth scripts as input. To enable support, pass "--enable-vapoursynth" to configure. Vapoursynth is detected via "pkg-config". Versions 42 or greater supported. See <http://www.vapoursynth.com/>.
Due to security concerns, Vapoursynth scripts will not be autodetected so the input format has to be forced. For ff* CLI tools, add "-f vapoursynth" before the input "-i yourscript.vpy".
FFmpeg can make use of the x264 library for H.264 encoding.
Go to <http://www.videolan.org/developers/x264.html> and follow the instructions for installing the library. Then pass "--enable-libx264" to configure to enable it.
FFmpeg can make use of the x265 library for HEVC encoding.
Go to <http://x265.org/developers.html> and follow the instructions for installing the library. Then pass "--enable-libx265" to configure to enable it.
FFmpeg can make use of the xavs library for AVS encoding.
Go to <http://xavs.sf.net/> and follow the instructions for installing the library. Then pass "--enable-libxavs" to configure to enable it.
FFmpeg can make use of the xavs2 library for AVS2-P2/IEEE1857.4 video encoding.
Go to <https://github.com/pkuvcl/xavs2> and follow the instructions for installing the library. Then pass "--enable-libxavs2" to configure to enable it.
ZVBI is a VBI decoding library which can be used by FFmpeg to decode DVB teletext pages and DVB teletext subtitles.
Go to <http://sourceforge.net/projects/zapping/> and follow the instructions for installing the library. Then pass "--enable-libzvbi" to configure to enable it.
You can use the "-formats" and "-codecs" options to have an exhaustive list.
FFmpeg supports the following file formats through the "libavformat" library:
@tab 4X Technologies format, used in some games.
@tab Audible Enhanced Audio format, used in audiobooks.
@tab Audible Format 2, 3, and 4, used in audiobooks.
@tab contains G.729 audio
@tab Multimedia format used in games like Mad Dog McCree.
@tab Multimedia format used in game Heart Of Darkness.
@tab Audio only format used in some Interplay games.
@tab Audio format used on the Nintendo Gamecube.
@tab Audio format used on the Nintendo Gamecube.
@tab Audio format used on the PS2.
@tab Advanced / Active Streaming Format.
@tab Audio format used on the Nintendo Wii.
@tab Audio format used on Mac.
@tab Multimedia format used by the Creature Shock game.
@tab Audio and video format used in some games by Beam Software.
@tab Used in some games from Bethesda Softworks.
@tab Multimedia format used by many games.
@tab Audio only multimedia format used by some games.
@tab Used in Z and Z95 games.
@tab Argonaut Games format.
@tab Used in the game Flash Traffic: City of Angels.
@tab Audio format used on the Nintendo WiiU (based on BRSTM).
@tab Audio format used on the Nintendo Wii.
@tab Broadcast Wave 64bit.
@tab Must be given -mode format option to decode correctly.
@tab Contains header with version and mode info, simplifying playback.
@tab Audio-only format used in console video games.
@tab Audio-only format used in console video games.
@tab Used in the game Cyberia from Interplay.
@tab Multimedia format used by Delphine Software games.
@tab Video format used by CD+G karaoke disks
@tab Amiga CD video format
@tab Apple Core Audio Format
@tab Created for the Sound Blaster Pro.
@tab Audio format used in some games by CRYO Interactive Entertainment.
@tab This format is used in Chronomaster game
@tab This format is used in the non-Windows version of the Feeble Files
game and different game cutscenes repacked for use with ScummVM.
@tab Used in various EA games; files have extensions like WVE and UV2.
@tab Only embedded audio is decoded.
@tab .fli/.flc files
@tab Macromedia Flash video files
@tab Audio format used in various games from FunCom like The Longest Journey.
@tab Audio format for various games.
@tab General eXchange Format SMPTE 360M, used by Thomson Grass Valley
playout servers.
@tab Only version 4 supported, used in some games from Cryo Interactive
@tab Microsoft Windows ICO
@tab Used in Quake III, Jedi Knight 2 and other computer games.
@tab Interchange File Format
@tab A format used by some old CCTV DVRs.
@tab Format used in various Interplay computer games.
@tab I-frames only
@tab A format generated by IndigoVision 8000 video server.
@tab A format used by libvpx
@tab Limitless Audio Format
@tab Used by Linux Media Labs MPEG-4 PCI boards
@tab contains LATM multiplexed AAC audio
@tab VR native stream format, used by Leitch/Harris' video servers.
@tab Metadata in text format.
@tab Used in Sim City 3000; file extension .xa.
@tab Used in some games from Capcom; file extension .mca.
@tab Used by Megalux Ultimate Paint
@tab 3GP, 3GP2, PSP, iPod variants supported
@tab muxed audio and video, VCD format supported
@tab also known as C<VOB> file, SVCD and DVD format supported
@tab also known as DVB Transport Stream
@tab MPEG-4 is a variant of QuickTime.
@tab Audio format used on the PS3.
@tab No cursor rendering.
@tab Used by MSN Messenger webcam streams.
@tab SMPTE 377M, used by D-Cinema, broadcast industry.
@tab SMPTE 386M, D-10/IMX Mapping.
@tab NC (AVIP NC4600) camera streams
@tab Nippon Telegraph and Telephone Corporation TwinVQ.
@tab NUT Open Container Format
@tab Used by TechnoTrend DVB PCI boards.
@tab File format used by RED Digital cameras, contains JPEG 2000 frames and PCM audio.
@tab Audio and video format used in some games by Entertainment Software Partners.
@tab Output is performed by publishing stream to RTMP server
@tab Used in many Sega Saturn console games.
@tab .sol files used in Sierra Online games.
@tab Used in Sierra CD-ROM games.
@tab Multimedia format used by many games.
@tab Used in certain Loki game ports.
@tab Multimedia format used in some LucasArts games.
@tab Audio format used in Sony Sonic Stage and Sony Vegas.
@tab Audio format used in Konami PS2 games.
@tab Used on the Nintendo GameCube.
@tab Tiertex .seq files used in the DOS CD-ROM version of the game Flashback.
@tab Audio format used in many Sony PS2 games.
@tab Audio format used in Sony PS games.
@tab Multimedia format used in Origin's Wing Commander III computer game.
@tab Multimedia format used in Westwood Studios games.
@tab Multimedia format used in Westwood Studios games.
@tab Microsoft video container used in Xbox games.
@tab Audio format used on the PS3.
@tab Microsoft audio container used by XAudio 2.
"X" means that the feature in that column (encoding / decoding) is supported.
FFmpeg can read and write images for each frame of a video sequence. The following image formats are supported:
@tab one raw file per component
@tab Alias/Wavefront PIX image format
@tab Animated Portable Network Graphics
@tab Microsoft BMP image
@tab Argonaut BRender 3D engine image format.
@tab Cintel RAW
@tab Digital Picture Exchange
@tab OpenEXR
@tab Flexible Image Transport System
@tab Radiance HDR RGBE Image format
@tab GEM Raster image
@tab Progressive JPEG is not supported.
@tab Lossless JPEG
@tab Microsoft Paint image
@tab PAM is a PNM extension with alpha support.
@tab Portable BitMap image
@tab PhotoCD
@tab PC Paintbrush
@tab Portable FloatMap image
@tab Portable GrayMap image
@tab PGM with U and V components in YUV 4:2:0
@tab PGX file decoder
@tab Portable HalfFloatMap image
@tab Pictor/PC Paint
@tab Portable Network Graphics image
@tab Portable PixelMap image
@tab Photoshop
@tab V.Flash PTX format
@tab Quite OK Image format
@tab SGI RGB image format
@tab Sun RAS image format
@tab YUV, JPEG and some extension is not supported yet.
@tab Targa (.TGA) image format
@tab Vizrt Binary Image format
@tab Wireless Application Protocol Bitmap image format
@tab WebP image format, encoding supported through external library libwebp
@tab X BitMap image format
@tab X-Face image format
@tab X PixMap image format
@tab X Window Dump image format
"X" means that the feature in that column (encoding / decoding) is supported.
"E" means that support is provided through an external library.
@tab Used in certain computer games.
@tab Creates video suitable to be played on a commodore 64 (multicolor mode).
@tab Used in games like Mad Dog McCree.
@tab Used in Chinese MP3 players.
@tab fourcc: apch,apcn,apcs,apco,ap4h,ap4x
@tab fourcc: qdrw
@tab Used in some Argonaut games.
@tab fourcc: ASV1
@tab fourcc: ASV2
@tab fourcc: VCR1
@tab fourcc: VCR2
@tab fourcc: AASC
@tab Supported through external libraries libaom, libdav1d, librav1e and libsvtav1
@tab fourcc: AVrp
@tab Video encoding used by the Creature Shock game.
@tab Supported through external libraries libxavs2 and libdavs2
@tab Supported through external library libuavs3d
@tab Microsoft uncompressed packed 4:4:4:4
@tab Used in some games from Bethesda Softworks.
@tab fourcc: BT20
@tab Used in the game Flash Traffic: City of Angels.
@tab Codec used in Cyberia game.
@tab fourcc: CSCD
@tab Video codec for CD+G karaoke disks
@tab Amiga CD video codec
@tab AVS1-P2, JiZhun profile, encoding through external library libxavs
@tab Codec used in Delphine Software International games.
@tab Codec used in various Broderbund games.
@tab fourcc: CLJR
@tab Codec used in Chronomaster game.
@tab supported though the native vc2 (Dirac Pro) encoder
@tab aka SMPTE VC3
@tab fourcc: DUCK
@tab fourcc: TM20
@tab fourcc: TR20
@tab Codec originally used in Feeble Files game.
@tab Used in NHL 95 game.
@tab lossless codec (fourcc: FFV1)
@tab fourcc: FSV1
@tab Sorenson H.263 used in Flash
@tab fourcc: G2M2, G2M3
@tab fourcc: G2M4
@tab encoding supported through external library libx264 and OpenH264
@tab encoding supported through external library libx265 and libkvazaar
@tab fourcc: ULTI
@tab Used in Quake II.
@tab Used in Quake III, Jedi Knight 2, other computer games.
@tab IFF interleaved bitmap
@tab IFF run length encoded bitmap
@tab Used in the game Cyberia from Interplay.
@tab Used in Interplay .MVE files.
@tab Codec used in Worms games.
@tab Kega emulator screen capture codec.
@tab Used in LucasArts games / SMUSH animations.
@tab Also known as Microsoft Screen 3.
@tab Also known as Microsoft Titanium Screen 2.
@tab Also known as Windows Media Video V7 Screen.
@tab Also known as Windows Media Video V9 Screen.
@tab Used in MSN Messenger Webcam streams.
@tab fourcc: VIXL
@tab libxvidcore can be used alternatively for encoding.
@tab Video encoding used in NuppelVideo files.
@tab still experimental
@tab fourcc: VP40
@tab fourcc: VP50
@tab fourcc: VP60,VP61,VP62
@tab fourcc: VP70,VP71
@tab fourcc: VP80, encoding supported through external library libvpx
@tab encoding supported through external library libvpx
@tab fourcc: Y216
@tab fourccs: QPEG, Q1.0, Q1.1
@tab fourcc: 'rle '
@tab fourcc: 'smc '
@tab fourcc: rpza
@tab still far from ideal
@tab Texture dictionaries used by the Renderware Engine.
@tab fourcc: 'RTV1'
@tab used in some games by Entertainment Software Partners
@tab Used in Sierra VMD files.
@tab Video encoding used in Smacker.
@tab experimental wavelet codec (fourcc: SNOW)
@tab fourcc: SVQ1
@tab fourcc: SVQ3
@tab fourcc: SP5X
@tab fourcc: TSCC
@tab fourcc: TSC2
@tab encoding supported through external library libtheora
@tab Codec used in DOS CD-ROM FlashBack game.
@tab fourcc: 'VMX1'
@tab Codec used in videos captured by VMware.
@tab not completely working
@tab Used in Wing Commander III .MVE files.
@tab Used in Wing Commander IV.
@tab libquicktime uncompressed packed 4:2:0
@tab part of LCL, encoder experimental
@tab Encoder works only in PAL8.
"X" means that the feature in that column (encoding / decoding) is supported.
"E" means that support is provided through an external library.
@tab encoding supported through internal encoder and external library libfdk-aac
@tab encoding supported through external library libfdk-aac
@tab 16 -E<gt> 4, 8 -E<gt> 4, 8 -E<gt> 3, 8 -E<gt> 2
@tab Used in various EA titles.
@tab Used in Sim City 3000.
@tab Used in AMV files
@tab Used in FunCom games.
@tab Used in some Sega Saturn console games.
@tab Used in some Sega Saturn console games.
@tab Used in Sega Dreamcast games.
@tab Used in LucasArts SMUSH animations.
@tab Used in Westwood Studios games like Command and Conquer.
@tab encoding supported through external library libopencore-amrnb
@tab encoding supported through external library libvo-amrwbenc
@tab QuickTime fourcc 'alac'
@tab Used in Bluetooth A2DP
@tab Used in Bluetooth A2DP
@tab Used in Bink and Smacker files in many games.
@tab decoding supported through external library libcelt
@tab en/decoding supported through external library libcodec2
@tab Codec used in Delphine Software International games.
@tab All versions except 5.1 are supported.
@tab supported extensions: XCh, XXCH, X96, XBR, XLL, LBR (partially)
@tab Used in few games.
@tab Used in Quake III, Jedi Knight 2 and other computer games.
@tab Used in various Interplay computer games.
@tab Used in various games.
@tab Used in Sierra Online game audio files.
@tab Used in Origin's Wing Commander IV AVI files.
@tab encoding supported through external library libgsm
@tab encoding supported through external library libgsm
@tab encoding and decoding supported through external library libilbc
@tab Used in DVD-Audio discs.
@tab encoding supported also through external library TwoLAME
@tab encoding supported through external library LAME, ADU MP3 and MP3onMP4 also supported
@tab encoding supported through external library libopus
@tab There are still some distortions.
@tab Real 14400 bit/s codec
@tab Real 28800 bit/s codec
@tab Real low bitrate AC-3 codec
@tab Used in Bluetooth A2DP
@tab Used in Sierra VMD files.
@tab experimental codec
@tab experimental codec
@tab supported through external library libspeex
@tab Used in HD-DVD and Blu-Ray discs.
@tab Used in LucasArts SMUSH animations.
@tab A native but very primitive encoder exists.
"X" means that the feature in that column (encoding / decoding) is supported.
"E" means that support is provided through an external library.
"I" means that an integer-only version is available, too (ensures high performance on systems without hardware floating point support).
"X" means that the feature is supported.
"E" means that support is provided through an external library.
"X" means that the protocol is supported.
"E" means that support is provided through an external library.
"X" means that input/output is supported.
ffplay(1), ffmpeg(1), ffprobe(1), ffmpeg-utils(1), ffmpeg-scaler(1), ffmpeg-resampler(1), ffmpeg-codecs(1), ffmpeg-bitstream-filters(1), ffmpeg-formats(1), ffmpeg-devices(1), ffmpeg-protocols(1), ffmpeg-filters(1)
The FFmpeg developers.
For details about the authorship, see the Git history of the project (https://git.ffmpeg.org/ffmpeg), e.g. by typing the command git log in the FFmpeg source directory, or browsing the online repository at <https://git.ffmpeg.org/ffmpeg>.
Maintainers for the specific components are listed in the file MAINTAINERS in the source code tree.